1. Introduction
Distyrylbenzenes are photoactive compounds applicable for the design of organic light-emitting diodes, solar cells, nonlinear-optical materials, and chemical sensors [
1,
2,
3,
4,
5,
6]. Owing to the presence of two C=C double bonds, distyrylbenzene molecules are characterized by light sensitivity, which provides additional possibilities for their phototransformations. Distyrylbenzenes are addressed in quite a few publications [
7,
8,
9,
10,
11,
12,
13,
14,
15,
16], but no data on the study of photoprocesses in distyrylbenzenes and their derivatives are available. Previously, we investigated the photonics of 1,4- and 1,3-diazadistyrylbenzene derivatives containing quaternary nitrogen atoms as strong electron-withdrawing moieties [
17] and showed that the photoprocesses characteristics of the 1,4- and 1,3-diazadistyrylbenzene isomers are considerably different.
The phototransformations of 1,4-diazadistyrylbenzene derivatives include
trans–
cis isomerization and the intersystem crossing to the triplet state. In this case, the formation of dihydrophenanthrene (DHP) is energetically unfavorable. For 1,3-diazadistyrylbenzene, the
trans–cis isomerization is followed by electrocyclization of the
cis–form to DHP and the subsequent fast oxidation with oxygen to give phenanthrene, whereas the intersystem crossing to the triplet state is not effective. This study is devoted to 1,4- and 1,3-distyrylbenzene derivatives containing strong electron-donating diethylamino groups. The structures with atom numbering differing from the IUPAC rules of the studied distyrylbenzenes are shown in
Figure 1.
The purpose of this study is to compare the photonics of compounds in MeCN using absorption, fluorescence, and laser kinetic spectroscopy, and to establish the presence of the triplet state and intermediate products of photoreactions. The experiments are accompanied by quantum chemical calculations in order to determine the chromophore structures in the ground and excited states and to interpret the experimental absorption and fluorescence spectra. 1H NMR spectroscopy was used to study the trans–cis photoisomerization of diethylamino 1,4- and 1,3-distyrylbenzenes.
2. Results and Discussion
2.1. Electronic Absorption and Fluorescence Spectra
Figure 2 shows the absorption spectra of compounds
1 and
2. The spectra of
1 and
2 differ in both the shape and positions of the absorption maxima. The absorption band of
2 is considerably blue-shifted (by 40 nm) compared to that of
1. The experimental and calculated absorption maxima are given in
Table 1.
Figure 3 shows the fluorescence spectra of
1 and
2, while
Table 2 summarizes the absorption maxima, Stokes shifts, and fluorescence quantum yields.
Fluorescence is observed for both 1 and 2. The fluorescence excitation spectra coincide with the absorption spectra of these compounds. For compound 2, there is a large Stokes shift of the fluorescence band (100 nm) and a considerable (almost 2.5-fold) decrease in the fluorescence quantum yield compared to 1; this may be attributable to a lower degree of conjugation in 2. Both compounds show P-type delayed fluorescence, indicating the intersystem crossing to the triplet state. The presence of P-type fluorescence is evidenced by the lack of temperature dependence of the delayed fluorescence intensity.
2.2. Kinetic Spectroscopy
Pulsed laser excitation of compound
1 results in a change in the absorption spectrum, as shown in
Figure 4.
The difference spectrum exhibits a decrease in the absorbance in the main absorption band of the dye (350–440 nm) and an increase in the absorbance in the 450–800 nm range with a maximum at 640 nm as a result of the formation of short-lived photoreaction products.
Figure 5 presents the corresponding kinetic curves measured at the photoinduced absorption maxima.
The kinetic curves in
Figure 5a demonstrate the effect of oxygen on the photoinduced absorption and attest to the presence of the intersystem crossing to the triplet state. The kinetics of the absorbance at λ = 630 nm is biexponential and is characterized by rate constants k
1 = 9 × 10
4 and k
2 = 6 × 10
3 s
−1. The absorbances at λ = 630 nm and 440 nm vary in a similar way (
Figure 5b), which attests to the formation of a photoproduct from the triplet excited state. In order to elucidate the nature of this photoproduct, we studied the effects of electron donors and acceptors on the kinetics of the phototransformations of
1.
Figure 6 and
Figure 7 show the kinetic curves for absorption measured in the absence of, and in the presence of an electron acceptor [
para-nitroacetophenone (PNAP)] and a donor [ascorbic acid (AA)], respectively. The effects of PNAP and AA on the kinetics of the photoinduced absorption lead to the conclusion that electron transfer occurs, producing the radical cation and the radical anion of compound
1, respectively.
In the presence of PNAP and AA, the time for the relatively fast changes in the induced absorption decreases, while the time for the relatively slow decrease in the induced absorption at λ = 630 nm increases, along with a similar increase in the absorption at λ = 440 nm.
The observed results may be interpreted by assuming that the electron transfer, which results in the formation of radical species, occurs either due to the dismutation of molecules in the triplet state or due to the reaction of the triplet molecules with the dye molecules in the ground state [
18,
19,
20]:
The fast decrease in the absorbance at 630 nm can be attributed to the R
•− radical, whose lifetime is reduced in the presence of oxygen, while the relatively slow stage can be associated with the lifetime of the R
•+ radical. The increase in the duration of photoproduct accumulation at 440 nm, accompanied by a parallel decrease in the absorption at 630 nm, is attributable to the formation of radical cations and radical anions in the presence of an electron donor (D) and acceptor (A) when they react with the dye molecule in the triplet state [
20,
21].
Thus, in the presence of high concentrations of an electron donor or acceptor, reaction (1) or (2) is suppressed, while reaction (3) or (4) takes place. The observed photochemical transformations attest to the similarity between the absorption spectra of the dye radical cations and radical anions.
The pulsed laser excitation of compound
2 induces a change in the absorption spectrum depicted in
Figure 8.
The difference spectrum shows an increased absorbance over the whole spectral range as a result of the formation of short-lived photoreaction products. The absence of a decrease in the absorbance at the major absorption band of the dye can be explained by the formation of a photoproduct with an absorption spectrum similar to that of 2, which may correspond to the cis-isomer of the dye.
Figure 9 shows the corresponding kinetic curves measured at the photoinduced absorption maxima. The effect of air oxygen on the photoinduced absorption kinetics supports the conclusion that intersystem crossing to the triplet state of the dye occurs.
The kinetic curves presented in
Figure 10 indicate a similar variation of the induced absorption at λ = 460 and 360 nm.
Similar to compound
1, we can assume that T–T dismutation reaction takes place to give radical cations and radical anions of the dye (Reaction (1)). The effects of the electron donor (AA) and acceptor (PNAP) on the photoreaction kinetics are depicted in
Figure 11 and
Figure 12.
Similar to compound 1, the experimental results can be interpreted by assuming that an electron acceptor or electron donor, when present in high concentration, reacts with the triplet molecule of dye 2 (Reactions (2) and (3)), thereby preventing the T–T dismutation (Reaction (1)).
2.3. 1H NMR Spectra
The
1H NMR spectra of distyrylbenzenes
2 and
1 (
Figure 13a and
Figure 14a) show one set of signals for each type of olefinic protons at 7.09–7.18 ppm for the β-H protons and at 6.93–6.98 ppm for the α-H protons, with spin–spin coupling constants of 16.3–16.4 Hz. These signals correspond to the (
E,
E)-isomers (proton numbering differing from the IUPAC rules is shown in
Figure 13 and
Figure 14). Data from the
1H NMR spectra indicate that distyrylbenzenes
1 and
2 were formed as (
E,
E)-isomers. After irradiation, the
1H NMR spectrum of 1,3-distyrylbenzene
2 (
Figure 13b) exhibits three sets of signals for aliphatic, olefinic α-H and β-H protons, along with most aromatic protons, which indicates the formation of a mixture of (
E,
Z)- and (
E,
E)-isomers. In general, the proton chemical shifts of the
cis-isomer are lower than those of the
trans-isomer. The aliphatic part of the spectrum provides little information, as proton signals for the (
E,
Z)-isomer appear as multiplets and are superimposed on the signals of the (
E,
E)-isomer. The
cis- and
trans-isomers of unsaturated compounds can be unambiguously identified by their spin–spin coupling constants, which are much smaller for
cis-interactions than for the
trans-isomers. The α-H and β-H signals for the (
E,
Z)-isomer are observed more upfield. Indeed, α-H protons give rise to three doublets: 6.42 ppm (
J = 12.2 Hz), 6.87 ppm (
J = 16.3 Hz), both for the (
E,
Z)-isomer and an intense doublet at 6.97 ppm (
J = 16.3 Hz) for the (
E,
E)-isomer. A similar pattern is observed for β-H protons, which give rise to three doublets: 6.53 ppm (
J = 12.2 Hz), 6.95 ppm (
J = 16.3 Hz), both for the (
E,
Z)-isomer, and an intense doublet at 7.17 ppm (
J = 16.3 Hz) for the (
E,
E)-isomer. The 3′-H and 5′-H aromatic protons also appear as three doublets: 6.57 and 6.71 ppm for the (
E,
Z)-isomer and an intense doublet at 6.73 ppm for the (
E,
E)-isomer. The 2′-H and 6′-H protons give rise to a doublet at 7.12 ppm for the (
E,
Z)-isomer. The second signal corresponding to the (
E,
Z)-isomer cannot be unambiguously identified, as it is located more downfield as part of a complex multiplet formed by the signals of the 4-H, 5-H, and 6-H aromatic protons of the central benzene ring for both isomers. The third intense doublet corresponding to the 2′-H and 6′-H protons of the (
E,
E)-isomer occurs at 7.42 ppm. The 2-H aromatic protons appear as two singlets at 7.47 and 7.66 ppm for (
E,
Z)- and (
E,
E)-isomers, respectively. Analysis of the integral intensities of proton signals corresponding to the (
E,
E)- and (
E,
Z)-isomers leads to the conclusion that the (
E,
Z)-isomer content in the mixture is at least 30%.
The
1H NMR spectrum of 1,4-distyrylbenzene
1 (
Figure 14b) recorded after irradiation is similar, but the intensity of the proton signals corresponding to the (
E,
Z)-isomer is much lower, and most of these signals are overlapped by those of the (
E,
E)-isomer. Two doublets for the α-H olefinic protons in the (
E,
Z)-isomer were unambiguously identified; they appear more upfield, at 6.37 ppm (
J = 11.9 Hz) and 6.92 ppm (
J = 16.2 Hz), while the intense doublet corresponding to the α-H proton in the (
E,
E)-isomer (
J = 16.4 Hz) is at 6.94 ppm. Analysis of the integral intensities of proton signals corresponding to the (
E,
E)- and (
E,
Z)-isomers indicates that the content of the (
E,Z)-isomer in the mixture does not exceed 10%.
Based on the 1H NMR data, we can conclude that isomerization involves one double bond and that the mixture contains both (E,E)- and (E,Z)-isomers. Distyrylbenzene 1 is less prone to E-Z photoisomerization, which is to be expected because the degree of conjugation between molecular groups is higher for the para-substituted derivative.
2.4. Quantum Chemical Calculations
The objectives of the quantum chemical calculation are (1) to calculate the energy differences of compounds
1 and
2 and their conformers in order to estimate the conformational composition, (2) to calculate λ
max in the absorption and fluorescence spectra, and (3) to elucidate, at the orbital levels the differences in the chromophore properties of these compounds.
Table 3 illustrates the structures of the possible isomers and conformers (rotamers) of compounds
1 and
2 calculated with full opgeometry optimization at the B3LYP/DFT/D3BJ level of theory. The possible isomers include
trans-
cis (
E–
Z)-isomers relative to the double C=C bonds. The terms “
syn” and “
anti” stand for the conformers relative to the σ-bonds between the substituents and the central benzene ring. For the
Z-isomers, only the mono-
Z-isomers were calculated, since only one double bond was experimentally found to be isomerized upon irradiation of the initial trans forms
1 and
2. The structures shown in
Table 3 are two-dimensional projections constructed using calculated Cartesian atomic coordinates and provide a visual representation of the changing shape of the molecules.
Table 4 summarizes the energy and spectral parameters of the most favorable conformers and isomers.
It can be seen from
Table 4 that the
trans-forms of
1 and
2 in the ground S
0 and excited S
1 states are energetically the most favorable, with the energy differences between the
trans-
1 and
trans-
2 rotamers being low (0.13 and 0.04 kcal/mol, respectively) in comparison with
kT. The relative contents of rotamers were calculated by the formula, as follows [
22]:
where Δ
E is the energy difference between the conformers,
k is the Boltzmann constant, and
T is the temperature (293 °C). This indicates that the presence of both conformers of
trans-
1 in the equilibrium mixture amounts to 55% (in favor of
anti,
anti-(
E,
E)-
1) and 45% for the other conformer, meaning that they are present in approximately equal amounts. As regards compound
2, the energy difference between the conformers is 3.46 kcal/mol (
Table 4). This is sufficient for
anti,
anti-(
E,
E)-
2 to predominate in the reaction mixture (99.7%). In the
1H NMR spectra discussed above, we observed the averaged proton signals for all possible conformers present in the solution.
The results of the quantum chemical calculations of spectral parameters are in good agreement with the spectroscopic data. However, since TDDFT calculations do not result in the exact agreement between the calculated and experimental absorption and fluorescence λ
max values [
23], we will discuss the trends in these values when moving from
1 to
2 and from
trans- to
cis-configuration. The calculation reproduced the blue shift of λ
max in the absorption spectra observed when moving from
1 to
2, although the absolute values of calculated λ
max are somewhat overestimated. The blue shift of λ
max is due to the different orbital structures of
1 and
2, caused by the different types of conjugation in
para- and
meta-isomers. The structures of the frontier MOs are shown in
Figure 15.
Figure 15 demonstrates how the difference between the
para- and
meta-substitution is manifested. In the case of
syn,
syn-(
E,
E)-
1, excitation involves an electron transition from the highest occupied MO (HOMO), which delocalizes over the whole molecule, to the lowest unoccupied MO (LUMO), the electron density maximum of which is concentrated mainly in the central part of the molecule and, to a lesser extent, on the dimethylaniline moieties. A different situation is observed for
anti,
anti-(
E,
E)-
2. In this case, excitation induces migration of an electron from HOMO-1, the electron density maximum of which is concentrated in the right-hand part of the molecule and does not affect five C atoms of the central benzene ring, to LUMO, the electron density maximum of which is also concentrated in the central and left-hand parts. However, the main difference between the orbitals of
1 and
2 is that in the latter compound, there are nodal planes of HOMO-1 that pass through 1–6 and 3–4 C atoms of the central benzene ring and a LUMO nodal plane passing through the 2–5 C atoms of
2. This means that the right and left halves of the molecule are isolated from each other. It is worth noting that in reality, this isolation leads to a partial loss of conjugation, manifested in a blue shift of the absorption band and estimated as 1.1 kcal/mol.
The calculation shows that the transition from trans- to cis-forms of 1 and 2 is accompanied by a blue shift of absorption bands (nm): from 425 to 408 and from 397 to 382, respectively.
3. Materials and Methods
Diethylamino 1,4- and 1,3-distyrylbenzenes (
1,
2) have been previously reported in the literature [
24,
25], but here they were synthesized by original procedures that will be published elsewhere. The structure and the purity of compounds
1 and
2 were confirmed by
1H NMR spectra. The structures with atom numbering differing from the IUPAC rules are shown in
Figure 1.
1H NMR spectrum of
1 (
Figure S1) and
2 (
Figure S2) in MeCN are available in the
Supplementary Material of this article.
The 1H NMR spectra were recorded on a Bruker (Billerica, MA, USA) DRX-500 spectrometer (500.13 MHz) in MeCN-d3 at 25 °C using residual solvent protons as the internal standard (δH 1.96 ppm). The chemical shifts were measured with an accuracy of 0.01 ppm, while the accuracy of measurement of the spin–spin coupling constant was 0.1 Hz.
Melting points were measured in capillaries on a Mel-Temp II device.
(
E,
E)-1,4-Bis(diethylaminodistyryl)benzene (
1).
Tm = 242–244 °C (cf. lit.: 242–246 °C) [
16].
1H NMR (500 MHz, MeCN-
d3), 25 °C, δ: 1.16 (t, 12H,
J = 7.0 Hz, 4
MeCH
2), 3.41 (q, 8H,
J = 7.0 Hz, 4
CH2Me), 6.72 (d, 4H,
J = 8.8 Hz, 2 3′-H, 2 5′-H), 6.94 (d, 2H,
J = 16.4 Hz, 2 α-H), 7.11 (d, 2H,
J = 16.4 Hz, 2 β-H), 7.41 (d, 4H,
J = 8.8 Hz,2 2′-H, 2 6′-H), 7.48 (br s, 4H, 2-H, 3-H, 5-H, 6-H).
(
E,
E)-1,3-Bis(diethylaminodistyryl)benzene (
2).
Tm = 136 °C. (cf. lit.: 134 °C) [
17].
1H NMR (500 MHz, MeCN-d3) 25 °C, δ: 1.17 (t, 12H, J = 7.0 Hz, 4MeCH2), 3.42 (q, 8H, J = 7.0 Hz, 4CH2Me), 6.73 (d, 4H, J = 8.9 Hz, 2 3′-H, 2 5′-H), 6.96 (d, 2H, J = 16.3 Hz, 2 α-H), 7.16 (d, 2H, J = 16.3 Hz, 2 β-H), 7.30–7.36 (m, 3H, 4-H, 5-H, 6-H), 7.42 (d, 2H, J = 8.9 Hz, 2 2′-H, 2 6′-H), 7.66 (s, 1H, 2-H).
The absorption spectra were recorded on an Agilent (Santa Clara, CA, USA) 8453 spectrophotometer. The luminescence spectra were measured on a Varian Eclipse spectrofluorometer (Agilent, USA). Delayed fluorescence and phosphoresce were recorded at 200 µs after the light pulse used for excitation which ruled out the contribution of the fast fluorescence to the total recorded emission signal. The difference absorption spectra of photoreaction products and their transformation kinetics were measured on a nanosecond laser photolysis unit [
23]. The irradiation was carried out with the 3rd harmonic of Nd-YAG laser (Solar) (
λ = 355 nm). Dissolved oxygen was removed from the solution by bubbling argon through it. The fluorescence quantum yields were determined using the 9,10-diphenylanthracene as the standard (0.95 quantum yield in ethanol) [
26]. The measurements were performed in MeCN (special purity grade, brand 0, Cryochrom, Saint Petersburg, Russia, Russia) at room temperature.
The 1H NMR experiment with irradiation of samples of compounds 1 and 2 was performed in the following way. First, the spectrum was recorded for a distyrylbenzene solution in MeCN-d3 prepared under red light, to prevent trans-cis isomerization. Then this solution was irradiated with unfiltered light of an F-SP-20-827-E27 fluorescent lamp for 30 min. After irradiation, the spectrum was recorded again.
Quantum chemical calculations for the isomers and conformers (rotamers) of molecules
1 and
2 in the ground and excited states were performed with full geometry optimization using B3LYP functional [
27] at the DFT/def2-TZVP and TDDFT/def2-TZVP level of theory within the ORCA 5.0.3 software package [
28]. The D3BJ dispersion correction [
29] was also taken into account. The results of the calculations are summarized in
Table 4.