Solvent Effects on Isomerization and Spectral Properties of

ARTICLE pubs.acs.org/JPCA

Solvent Effects on Isomerization and Spectral Properties of Photochromic-Switching Diarythene Derivatives in Polar and Apolar Solutions Suci Meng†,‡ and Jing Ma*,† †

Institute of Theoretical and Computational Chemistry, Key Laboratory of Mesoscopic Chemistry of MOE, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, 210093, People’s Republic of China ‡ School of Chemistry and Chemical Engineering, Jiangsu University, Zhenjiang, 212013, People’s Republic of China

bS Supporting Information ABSTRACT: The photocyclization behavior and dynamic conformational transition of photochromic switches of diarythene derivatives in solutions are investigated by using the density functional theory (DFT) and molecular dynamics (MD) simulations. Three possible conformations, antiparallel (anti), parallel (para), and twist, for the open-ring isomers of 1,2-bis(2-methylbenzothiophene-3-yl)maleic anhydride are located. Both PCM-B3LYP/6-31G* calculations and MD simulations demonstrate that anti and twist open-ring isomers can interconvert freely in n-hexane and acetonitrile solutions at room temperature. The statistical ratio of twist to anti isomers from MD simulations is 2.09 in n-hexane and 1.07 in CH3CN, in qualitative agreement with those (1.18 in n-hexane and 1.05 in CH3CN) estimated from Arrhenius analysis of DFT activation energies. The solvent polarity has little influence on the isomerization of open-ring isomers in the ground state. Due to the evident charge transfer upon excitations, the solvent effects on the electronic structures and absorption spectra of low-lying excited states (S1 and S2) are more significant. For such charge-transfer excited states, the long-range corrected functional CAM-B3LYP gives better agreement with the experimental spectra than B3LYP. The solvent polarity and polarization of the charge-transfer excited states are crucial for fabricating the novel functionalized photochromic molecular switches.

1. INTRODUCTION Photochromism has attracted considerable attention because of its potential application to molecular devices, such as optical memories and switches.1
5 Among the various photochromic systems, diarylethenes are the most promising candidates for these applications due to their fatigue-resistant and thermally irreversible properties.6
16 They were demonstrated to undergo a cyclization/cycloreversion electrocyclic reaction upon UV and visible irradiation (cf. Scheme 1). It was conceived that the openring isomers of diarylethenes in solutions have two conformations, antiparallel (anti) and parallel (para), in almost equal amounts.17 It was found that only the anti conformer can be converted to the closed-ring isomer, while the cyclization reaction is inhibited for the para isomer. This has been rationalized by the conservation of molecular orbital symmetry.18 Our calculations also demonstrated that the selected dithienylethenes, 1
8, have similar pictures of frontier orbitals, and the photocyclization reaction may proceed only for the anti conformer in a conrotatory mode (see subsection 3.1). In the present work, we locate another twist open-ring isomer from both density functional theory (DFT) calculations and molecular dynamics (MD) simulations. The ring closure/opening reactions of photochromic dithienylethenes have been studied theoretically in the gas phase and r 2011 American Chemical Society

on the gold surface through semiempirical and ab initio calculations.18
24 The CASSCF(10,10)/6-31G calculations showed that the energy barrier of the ring closure/opening process in the ground state was too high (about 50 kcal/mol) to achieve the on/off interconversion, but it might be easy to realize the cyclization/cycloreversion process in the low-lying excited states (the energy barriers are less than 10 kcal/mol).19 Moreover, it was found that the gold surface inhibited the ring closure process owing to the quenching of the excited state of the open conformation in the presence of the gold electrodes through semiempirical AM1 and ZINDO calculations.20 In addition, the light-induced dynamical switching was simulated by the quantum molecular dynamics, which displayed a swapping of the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) during a period of 100 fs.18 However, to the best of our knowledge, the solvent effects on the conformational interconversion and the excited-state electronic structures of open-ring isomers of dithienylethene derivatives and the optical properties in solutions are rarely explored.25

Received: June 6, 2011 Published: December 13, 2011 913

dx.doi.org/10.1021/jp210846b | J. Phys. Chem. A 2012, 116, 913–923

The Journal of Physical Chemistry A

ARTICLE

Scheme 1. Open- and Closed-Ring Isomers of Selected Dithienylethene Derivatives, 1
8

Figure 1. The torsional potential surfaces of the 4o molecule along the torsional angles of (a) θ1 and (b) θ2. The relative energies of the extrema of 4o were obtained from B3LYP/6-31G* calculations and PCFF results with PCFF, Mulliken, ESP, and NBO partial charges.

derivatives. In this implicit model, the solvent is characterized by a continuum dielectric constant, ε. The nonpolar n-hexane (ε = 1.89) and the polar acetonitrile (ε = 36.64) solvents were selected. Force field (FF)-based MD simulations with hundreds of discrete solvent molecules were also employed to investigate the dynamic conformational interconversion of the open-ring isomers as well as the local solute
solvent interactions in n-hexane and acetonitrile solutions. The DFT and MD simulations will demonstrate that three different conformations, anti, para, and twist, of open-ring isomers coexist in solutions. In comparison with the ground state, the influence of the solvent polarity on the excited-state electronic structure is more significant. In addition, the MD simulations present the detailed picture of the short-range solute
solvent interactions in nonpolar n-hexane and polar acetonitrile solutions. The knowledge of isomerization reaction pathways and the solvent effects on the electronic structures of the low-lying excited states of open-ring isomers is helpful for the modulation of the performance of photochromic switches.

The presence of the solvents with different polarity may affect the relative thermal stability as well as the ground- and excitedstate electronic structures of open-ring isomers.26 Two of the most studied dithienylethene compounds are 2,3-bis(2,4,5-trimethylthiophene-3-yl)maleic anhydride (3) and 1,2-bis(2-methylbenzothiophene-3-yl)maleic anhydride (4).27
34 The experimental measurements showed that the fluorescence spectra of the open-ring isomers in solution are highly dependent on the solvent polarity, while the absorption spectra were scarcely affected by the polarity of solvents.26,28 In the present work, we will try to understand the photochromic on/off behaviors of open-ring isomers 1
8 in solutions from theoretical calculations. The influence of solvent polarity on the relative thermodynamic stability, isomerization, and spectral properties of open-ring isomers of dithienylethene derivatives is exemplified by 4o, for which the experimental data are abundant. Both DFT calculations and MD simulations were applied to study the photocyclization, conformational interconversion, and solvent effects on electronic structures of the ground and lowlying excited states of open-ring isomers in solutions. Two solvent models, implicit and explicit, were employed. The polarized continuum model (PCM),35
43 one of the most widely used continuum dielectric method, was adopted to investigate the solvation effects on the electronic structures of the low-lying excited states and reaction pathways for isomerization of dithienylene

2. COMPUTATIONAL DETAILS 2.1. Isomerization of Open-Ring Isomers: DFT/PCM Calculations. To gain better understanding of the experimental obser-

vations, we calculated the open/closed-ring isomers of 1
8 with anti, twist, and para conformations at the B3LYP/6-31G* level. The C2 symmetry is restrained for anti and twist open/closedring isomers. There exist several different open-ring isomers, as shown in Figure 1. The torsional potential profile of the 914

dx.doi.org/10.1021/jp210846b |J. Phys. Chem. A 2012, 116, 913–923

The Journal of Physical Chemistry A

ARTICLE

Table 1. Relative Electronic Energies (kcal/mol) of Anti, Para, and Twist Isomers and the Corresponding Transition States (TS) of Isomerization in the Gas Phase and in Solutions using the B3LYP Functional at the 6-31G*, 6-31+G*, and 6-311++G** Levels, Respectively 6-31G*

6-31+G*

6-311++G**

gas twist TSAT

0.21 5.02

0.08 4.32

0.11 4.12

anti

0.44

0.48

0.38

TSAP

17.16

17.47

17.31

para

0.00

0.00

0.00

TS

5.66

4.76

4.63

twist

0.30

0.17

0.15

TSAT

4.83

4.15

3.96

anti TSAP

0.32 17.38

0.34 17.70

0.25 17.56

para

0.00

0.00

0.00

TS

5.37

4.49

4.34

twist

0.47

0.42

0.42

TSAT

4.32

3.71

3.51

anti

0.06

0.04


0.05

TSAP para

17.96 0.00

18.39 0.00

18.25 0.00

4.56

3.72

3.53

n-hexane

Figure 2. Absorption spectra of anti, twist, and para conformers obtained from TD-DFT calculations using (a) B3LYP and (b) CAM-B3LYP functionals at the 6-31G* level in the gas phase and solutions with increasing the solvent polarity.

CH3CN

TS

Table 2. Optimized Geometries of the Anti Conformation (C2 symmetry) of the 4o Molecule (Bond Length in Å; Dihedral Angel in Degree) through PCFF, PCFF with Mulliken Charges, and B3LYP/6-31G* Calculations, Respectively

interconversion of open-ring isomers 4o in the ground state was calculated at the B3LYP/6-31G* level (Figure 1). All of the stationary points were tested by the vibrational frequency calculations. To qualitatively assess the solvent effects on the electronic structures of 4o, the torsional potential surfaces in n-hexane and acetonitrile were also studied in the solvents by adopting the PCM. All of these calculations were carried out through the Gaussian 03 program.44 2.2. Choice of Basis Sets and DFT Functionals. To test the basis set dependence, the relative electronic energies of minima and the corresponding transition states of open-ring isomers were calculated using the B3LYP functional with 6-31G*, 6-31+G*, and 6-311++G** basis sets. Table 1 shows the qualitative agreement for the relative electronic energies of isomers by using different basis sets. The electronic energies of three enantiomers (anti, para, and twist) of open isomers are very close to each other, with energy difference of less than 0.5 kcal/mol, implying the facial isomerization. The B3LYP functional has been widely employed to calculate molecular geometries,45 HOMO/LUMO gaps,46 and TD-DFT excitation energies.47 However, the application of TD-B3LYP to the spectral properties was questioned presumably due to the insufficiency in describing the strong charge-transfer character.48 It has been demonstrated that a new hybrid exchange
correlation functional using the Coulomb-attenuating method, CAM-B3LYP, gave more reliable results than traditional B3LYP, especially for the charge-transfer excited states.48 Thus, the CAM-B3LYP functional was also performed to calculate the absorption spectra using the

PCFF

PCFF-Mulliken

B3LYP/6-31G*

C1
C2

1.404

1.405

1.376

C2
C3

1.509

1.509

1.464

C3
C4

1.325

1.327

1.362

C2
C7 C7
C8

1.426 1.418

1.427 1.417

1.457 1.407

C8
C9

1.389

1.389

1.388

C9
C10

1.408

1.408

1.405

C10
C11

1.397

1.397

1.390

C11
C12

1.381

1.383

1.397

S13
C1

1.746

1.753

1.754

RC1 3 3 3 C6 θ

3.731
56.5

3.781
58.3

3.750
49.5

Gaussian 09 package.49 Figure 2 shows a comparison of TD-DFT excitation energies for open-ring isomers of the 4o molecule between the use of B3LYP and CAM-B3LYP functionals. It is demonstrated that the TD-B3LYP calculations overestimate the maximum absorption peaks (anti: 514
526 nm; twist: 499
506 nm; and para: 501
510 nm from n-hexane to acetonitrile) of open-ring 915

dx.doi.org/10.1021/jp210846b |J. Phys. Chem. A 2012, 116, 913–923

The Journal of Physical Chemistry A

ARTICLE

Scheme 2. Frontier Orbitals Involved in thePhotochromic Cyclization/Cycloreversion Reaction of Diarylethene Derivatives (4)

isomers by about 100 nm relative to experimental values (404
40726 or 399
404 nm28), while the deviation in maximum absorption wavelengths between TD-CAM-B3LYP results (anti: 399
404 nm; twist: 395
396 nm; and para: 394
396 nm) and experimental data is only 8
13 nm upon going from n-hexane to acetonitrile. Moreover, the solvent shifts of absorption spectra, Δλsol max, obtained by TD-CAM-B3LYP calculations (1
5 nm) are closer to the experimental observations (3
5 nm)26,28 relative to the TD-B3LYP results (7
12 nm). 2.3. Ground and Excited States: TD-DFT/PCM Calculations. The geometries of ground state (S0) and low-lying singlet

excited states (S1 and S2) of 4o isomers were optimized by DFT and TD-DFT calculations with the framework of PCM at the B3LYP/6-31G* level. In order to further assess the influence of solvent polarity on the absorption spectra, TD-CAMB3LYP/PCM vertical excitation energies were calculated on the ground (S0) as well as first (S1) and second (S2) singlet excited states of 4o isomers in n-hexane and acetonitrile solutions, respectively. 2.4. Conformational Variations and Intermolecular Interactions in Solutions: MD Simulations. The investigation of the conformational changes of open-ring isomers in the explicit 916

dx.doi.org/10.1021/jp210846b |J. Phys. Chem. A 2012, 116, 913–923

The Journal of Physical Chemistry A

ARTICLE

Figure 3. The LUMO orbitals of open-ring isomers and HOMO orbitals of closed-ring isomers along with their energy levels (in units of au and at the B3LYP/6-31G* level) of dithienylethene derivatives, 1
8. These molecular orbitals are plotted with isocontour values of 0.04 au.

solvent model as well as the intermolecular interactions requires the long time MD simulations. The force in each MD step is obtained from molecular mechanics, whose performance depends on the proper selection of force fields. In this work, we adopted the polymer consistent force field (PCFF),50
53 which has been applied to describe the packing structures of π-conjugated systems in solutions54
56 and the amorphous phase57,58 as well as on the Ag(111) surface.59,60 To further survey the impact of partial charges on the torsional potentials, the relative energies of conformers along the torsional potential surface were calculated through B3LYP/6-31G* calculations

and PCFF calculations with the Mulliken, electrostatic potential (ESP), and natural bond orbital (NBO) charges, respectively. From Figure 1, it is seen that the PCFF with Mulliken charges reproduces well the B3LYP/6-31G* results. In addition, the geometries of anti conformers obtained by B3LYP/ 6-31G* and PCFF (with default charges) as well as PCFF with Mulliken charge optimizations are compared in Table 2. We can see that the PCFF geometries with default and Mulliken charges qualitatively agree with the B3LYP/6-31G* results. In the following MD simulations, the PCFFs with Mulliken partial charges are employed. 917

dx.doi.org/10.1021/jp210846b |J. Phys. Chem. A 2012, 116, 913–923

The Journal of Physical Chemistry A

ARTICLE

The MD simulations of ring-opening isomers of diarylethene 4o in n-hexane and acetonitrile solutions were performed in the canonical (NVT) ensemble at 298 K by using a Nos
e
Hoover thermostat.61 Equations of the motion for systems were integrated using the velocity Verlet algorithm62 with the time step of 1 fs. The period boundary condition (PBC) was employed. Herein, the concentration of solution was set to be 0.1 M to mimic the dilute solution, with the number of the solvent molecules taken as 73 (n-hexane) and 177 (acetonitrile). The cutoff of van der Waals interactions was set to 15.5 Å. The electrostatic interaction was evaluated by the Ewald summations.63 The 3 ns simulations were subsequently carried out after the equilibrium

stage. Trajectories were collected every 100 fs. Finally, we employed trajectories of the 3 ns for the statistical analysis of solvent and solute configurations. All MD simulations were carried out using the discover module in Materials Studio package.64

3. RESULTS AND DISCUSSION 3.1. Photocyclization: A Qualitative Picture. As mentioned before, the switching events of the photochromic molecule were understood by the conservation of molecular orbital symmetry.18 Our calculations also show that the irreducible representation of HOMO (H) and LUMO (L) orbitals of anti open/closed-ring isomers is A (symmetry) or B (antisymmetry), respectively. The LUMO orbitals of anti open-ring isomers (reactants) do correlate with the HOMO orbitals of the corresponding closed forms (products) with the same symmetry relative to C2 operation, as displayed in Scheme 2. In fact, frontier molecular orbital theory is also applicable to the cyclization reaction of open-ring isomers of dithienylethene derivatives. In the language of frontier orbital theory, the photocyclization reaction involves the phase properties of the LUMO orbitals of open-ring isomers. The B3LYP/ 6-31G* calculations demonstrate that the selected diarylethenes 1
8 have a similar topological picture of the frontier molecular orbitals (cf. Figure 3). The photocyclization reaction for the antiparallel conformer is favored in a conrotatory mode, as schematically illustrated in Figure 3. Moreover, the open-ring anti and para conformations of dithienylethene derivatives can also transform into the corresponding closed forms, as shown in Scheme 2. The photogenerated central C1
C6 bonds in para closed-ring isomers are weakened due to the repulsion of two methyl groups on the same side, as reflected by the difference of bond length of the closed-ring isomers (cf. Table 3). The central C
C bond length, RC1
C6 of the para (1.545
1.572 Å) closed-ring isomer is significantly longer than that of the anti (1.513
1.549 Å) closed forms. Accordingly, the energies of para closed-ring isomers are much higher (by about 12
15 kcal/mol) than those of anti

Table 3. Distance (RC1
C6) between Two Active Carbons and the Relative Electronic Energies of Anti and Para Closed Isomers of Dithienylethene Derivatives 1
8a RC1
C6 (Å) 1 2 3 4 5 6 7 8 a

ΔE (kcal/mol)

1c (anti)

1.513

0.00

1c (para)

1.545

12.15

2c (anti) 2c (para)

1.546 1.564

0.00 14.88

3c (anti)

1.544

0.00

3c (para)

1.564

13.06

4c (anti)

1.548

0.00

4c (para)

1.564

11.90

5c (anti)

1.546

0.00

5c (para)

1.567

13.72

6c (anti) 6c (para)

1.549 1.572

0.00 14.66

7c (anti)

1.546

0.00

7c (para)

1.567

14.08

8c (anti)

1.549

0.00

8c (para)

1.572

14.93

The data were obtained from B3LYP/6-31G* optimizations.

Figure 4. Ground-state potential surfaces for the interconversion among the open-ring isomers (4o) in the gas phase, n-hexane, and acetonitrile (CH3CN). 918

dx.doi.org/10.1021/jp210846b |J. Phys. Chem. A 2012, 116, 913–923

The Journal of Physical Chemistry A

ARTICLE

Figure 5. The plots of electrostatic potentials of (a) para and (b) twist open-ring isomers in the gas phase and in acetonitrile solution. The dipole moments are in units of Debye.

Figure 6. Absorption spectra of the ground state (S0) and low-lying excited states (S1 and S2) of (a) para and (b) twist conformers in the gas phase and solutions with increasing solvent polarity. The insets show the energy levels of HOMO and LUMO orbitals from the gas phase to acetonitrile solutions. The data were obtained from TD-CAM-B3LYP/ PCM and B3LYP/PCM calculations using 6-31G* basis set.

Figure 7. Probability distributions of atomic distances, RC1 3 3 3 C6 between two reactive carbons (C1 and C6) of open-ring isomers 4o (anti and twist isomers) in (a) n-hexane and (b) acetonitrile (CH3CN) solutions. The data come from the statistical analysis of MD simulations of anti and twist conformations as the initial model, respectively, in n-hexane and acetonitrile solutions.

closed isomers, which suggests that the para closed-ring conformations may be difficult observe in experiments. 3.2. Isomerization of Open-ring Ground-state Isomers: DFT/PCM Calculations. Figure 4 shows the conformational interconversion between three open-ring isomers, twist, anti, and para, which are obtained from the geometry optimizations in the gas phase and solutions at the B3LYP/6-31G* level. A series of X-ray crystal data of anti conformations for open-ring isomers have been reported in experiments.10 As displayed in Table S1 (Supporting Information), the optimized geometries of the 1,2-bis(2-methyl-5-phenyl-3-thienyl)perfluorocyclopentene molecule are in agreement with the experimental data obtained from

single-crystal X-ray diffraction.10b As illustrated in Figure 4, in the anti isomer, two terminal methyl groups on the thiophene rings are arranged in an antiparallel manner with a short C1 3 3 3 C6 distance (3.76 and 3.80 Å in n-hexane and acetonitrile, respectively), facilitating the ring closing. In the para conformer, two terminal methyl groups are rotated to the same side of the anhydride. In contrast, the thiophene-based groups are twisted in the opposite sense in the twist conformation. In comparison with the anti isomer, both para and twist conformations have longer C1 3 3 3 C6 distances (para: 4.41
4.42 Å; twist: 5.67
5.69 Å), implying the difficulty in the phothochemical ring-closure reaction. In addition, there are three enantiomers, anti0 , para0 , and 919

dx.doi.org/10.1021/jp210846b |J. Phys. Chem. A 2012, 116, 913–923

The Journal of Physical Chemistry A

ARTICLE

Figure 8. Probability distributions of the torsional angle, θ, of open-ring isomers for (a) anti and twist isomers as well as (b) para and para0 conformations in n-hexane and acetonitrile (CH3CN) solutions. The data come from the statistical analysis of MD simulations in n-hexane and acetonitrile solutions, respectively.

twist0 isomers, with the same energy as the anti, para, and twist conformations both in the gas phase and in solutions (cf. Figure 4). Furthermore, it is demonstrated that three open-ring isomers, anti, para, and twist, of the 4o molecule may coexist in solutions because the energy difference between each path of transformation is less than 0.5 kcal/mol, as schematically illustrated in Figure 4. A facile interconversion between anti and twist conformations is also shown with the low-energy barrier of 4.78
5.10 kcal/mol. The isomerization between para and para0 conformers may also take place easily with an energy barrier of about 4.79
5.14 kcal/mol. However, the much higher potential barrier (in the gas phase: 18.60 kcal/mol; in n-hexane: 18.96 kcal/mol; in acetonitrile: 19.53 kcal/mol) separates the anti and para conformations, suggesting the difficult interconversion between anti and para isomers at room temperature. In addition, the influence of the solvent polarity on the energy barriers of interconversion is insignificant. The solvent shifts in energy barriers (0.1
0.9 kcal/mol) are negligible upon going from the gas phase to acetonitrile. The solvent shifts in the thermal freeenergy difference of the open-ring isomers are also very small (less than 0.5 kcal/mol), as shown in Figure 4. It is not surprising because the magnitudes of dipole moments in the openring isomers change little with the solvent polarity. In n-hexane and acetonitrile solutions, the dipole moments of anti are 5.15 and 5.93 D, for para, they are 5.12 and 5.82 D, and for twist, they are 5.01 and 5.55 D, respectively. 3.3. Solvent Effects on Low-Lying Excited States: The Extent of Charge Transfer. The TD-DFT optimized geometries of para and twist isomers in ground (S0) and low-lying excited states (S1 and S2) from the gas phase to CH3CN solution are displayed in Table S2 and Figure S1 (Supporting Information). In comparison with the ground-state geometries, the parameter of bond length alternation (BLA) in the backbone (highlighted in Figure S1, Supporting Information) of the para and twist isomers is decreased in the low-lying excited (S1 and S2) states. It suggests that the electron delocalization of para and twist conformers is enhanced from the ground (S0) state to the first (S1) and second (S2) singlet excited states, facilitating the polarization in the atmosphere of solvents. The CdO double bonds of the para and twist open-ring isomers are polarized and lengthened to 1.220
1.232 Å in the S1 and S2 states from 1.200
1.203 Å in the S0 state. In addition, two independent torsional angles, θ1 and θ2, are significantly augmented by about 25
35° in the S1 state of the para isomer and about 25
28° in the S2 state of the twist conformer,

respectively, which decreases the coplanarity of the backbone of open-ring isomers. The obvious change of geometry structures of para and twist conformers results in the charge redistribution of open-ring isomers in the excited states. The Mulliken charges of the furan-2,5-dione ring become more negative (para: q1 =
0.68 to
0.86; twist: q1 =
0.60 to
0.82), as shown by the red region in the electrostatic potential (ESP), while the charges of two 2-methylbenzothiophene rings become more positive (q2 or q3, para: 0.69
0.86; twist: 0.30
0.41, labeled as blue in the ESP) in the S1 and S2 states, with respect to those (q1 =
0.26 to
0.28; q2 = 0.13
0.14; and q3 = 0.13
0.14) in the S0 state. The ESP images demonstrate that para and twist molecules are highly polarized in the S1 and S2 states, which results in a 3
4 fold increase in the magnitude of the dipole moment upon excitation, as displayed in Figure 5. That suggests the non-negligible solvent effects in the excited states. The increased extent of charge transfer in the excited states can be also seen from Figure 6, in which the simulated ground- (S0) and excited-state (S1 and S2) absorption spectra of para and twist isomers of 4o in the gas phase and solutions are schematically illustrated. Herein, a Lorentzian function65 is adopted with the spectral line width set to be 50 nm. It is found that the calculated absorption spectra undergo a red shift to different degrees as the solvent polarity increases. Moreover, the solvent effects on the excited-state electronic structures are significant, in agreement with the experimental observations.26,28 Table S3 (Supporting Information) shows the detailed results of the low-lying dipoleallowed vertical excitation energies (Eex) corresponding to the maximum absorption band (cf. Figures 2b and 6) and oscillator strengths (f) of open-ring isomers from the gas phase to CH3CN solution. All of the lowest dipole-allowed absorption bands of the open-ring isomers are mainly due to the π f π* transition from HOMO to LUMO excitation. From the insets in Figure 6, it is found that the red shifts of the absorption spectra mainly stem from the promotion of the HOMO (H) energy level from the gas phase to acetonitrile solution. The intramolecular charge transfer dominates the lowest vertical excitation energies of the 4o molecule in the gas phase and solutions, rationalizing the significant increase of the excited-state dipole moments. 3.4. Conformational Interconversion of Open-Ring Isomers: MD Simulations. As mentioned above, DFT/PCM calculations indicate that three open-ring isomers of the 4o molecule may coexist in solutions owing to the small energy differences (less than 0.50 kcal/mol). We then resort to MD simulations to 920

dx.doi.org/10.1021/jp210846b |J. Phys. Chem. A 2012, 116, 913–923

The Journal of Physical Chemistry A

ARTICLE

Figure 9. Local solute
solvent interactions of (a) O15 3 3 3 H, (b) O14 3 3 3 H, and (c) O16 3 3 3 H of anti and twist conformations in CH3CN solution. All of the data come from the statistical analysis of MD simulations of 4o in CH3CN solution.

Moreover, with the increase of the polarity from n-hexane to acetonitrile, the population of the anti conformation increases, while that of the twist conformation decreases, as shown in Figure 8a. It suggests that the polar solvent favors the stabilization of the anti conformations to some extent. The statistical results of MD simulations show that the ratio of twist to anti isomers is 2.09 in n-hexane and 1.07 in CH3CN, which is qualitatively in line with those (1.18 in n-hexane and 1.05 in CH3CN) estimated from Arrhenius analysis of DFT activation energies. In addition, the MD simulations also demonstrate facile interconversions between para and para0 conformations, with the same probability in n-hexane and acetonitrile solutions (cf. Figures 8b and S4, Supporting Information). The interconversion between anti and para isomers was not observed during our 3 ns MD simulations, in agreement with the QM predictions. The solution configurations of open-ring isomers of 4o in nonpolar n-hexane and polar acetonitrile solutions were also analyzed

investigate the dynamic conformational interconversion among the open-ring isomers in solutions. The conformational change of the open-ring isomers in explicit solvent molecules are detected for anti, twist, and para conformers in nonpolar n-hexane and polar acetonitrile. From the time evolutions of the dihedral angle, θ, and the C1 3 3 3 C6 distance, RC1 3 3 3 C6 (cf. Figures S2 and S3, Supporting Information), it is found that anti and twist conformations do interconvert to each other at room temperature. To investigate the influence of the initial conformations of open-ring isomers in molecular simulation, two independent 3 ns MD simulations, starting with the anti and twist initial configurations, respectively, were carried out in n-hexane and acetonitrile solutions. As illustrated in Figure 7, the statistical analyses of the C1 3 3 3 C6 distance, RC1 3 3 3 C6, have demonstrated that different initial conformations do not change the conclusions drawn from MD simulations. 921

dx.doi.org/10.1021/jp210846b |J. Phys. Chem. A 2012, 116, 913–923

The Journal of Physical Chemistry A from 3 ns MD trajectories. In Figure 9, the short-range solute
solvent interactions in solutions are revealed from radial distribution functions (RDF) and spatial distribution functions (SDF).66 We pay attention to the O 3 3 3 H intermolecular interaction between the oxygen atoms, O14, O15, and O16 of 4o and the hydrogen atoms of the solvents. The first solvation shell SDF for hydrogen atoms in acetonitrile solvent around three oxygen atoms, O14, O15, and O16, in the anti configuration of 4o is depicted by the gOpenMol package.67 The solvent molecules evenly distribute in a sphere, showing the isotropy in solute
solvent interactions. Within the first solvation shell, the average distances of RO 3 3 3 H are about 2.90
3.15 Å in n-hexane and acetonitrile solutions (cf. Figures 9 and S5, S6, and S7a, Supporting Information), which are larger than the sum of the van der Waals radii (H + O: 2.72 Å). The average O 3 3 3 H
C angles are distributed about 115
125° in n-hexane and 110° in acetonitrile (cf. Figures 9 and S7b, S8, and S9, Supporting Information). Thus, the solute
solvent O 3 3 3 H
C interactions are not significant in both nonpolar n-hexane and polar acetonitrile solutions.

4. CONCLUSIONS In this work, the photocyclization behavior of open-ring isomers of diarylethenes is rationalized through theoretical calculations. The isomerization and spectral properties of open-ring isomers of 1,2-bis(2-methylbenzothiophene-3-yl)maleic anhydride are investigated in n-hexane and acetonitrile solutions through quantum chemical calculations and MD simulations. It has been demonstrated that three open-ring isomers, anti, para, and twist conformations, may exist in solutions. Moreover, both the DFT calculations and MD simulations show that the anti and twist conformers can interconvert easily, but the interconversion between anti and para isomers is relatively difficult at room temperature in the gas phase and solutions. The polar solvent favors the stabilization of the anti conformer with photocyclization reactivity. The enantiomers, para and para0 , can interconvert freely with equal population in solutions. QM calculations display that the solvent effects on the isomerization of open-ring isomers and electronic structures are relatively weak. The MD simulations also demonstrate that the local solute
solvent interactions are insignificant in both nonpolar n-hexane and polar acetonitrile solutions. The significant solvatochromism of excited-state absorption spectra of open-ring isomers is attributed to the evident increase of the dipole moments and the intramolecular charge transfer upon excitations. This detailed information for the conformational interconversion, electronic structures, and absorption spectra of open-ring conformers may shed light on the modulation of photoelectric properties and stabilization of the optical reactive conformations in the design of novel photochromic switches.

ARTICLE

solvent effects on the geometries of ground and low-lying excited states of para and twist open-ring isomers (Figure S1); evolution of the torsional angle and distance of 4o as a function of the simulation time in n-hexane and acetonitrile solutions (Figures S2
S4); radial distribution functions for oxygen atoms in open-ring isomers of 4o to hydrogen atoms in solvents and probability distributions of the O 3 3 3 H
C angle of open-ring isomers in n-hexane and acetonitrile solutions (Figures S5
S9). This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ ACKNOWLEDGMENT This work was supported by the National Natural Science Foundation of China (Grants No. 20825312 and 21103073), the National Basic Research Program (Grant No. 2011CB808600), UJS (Grant No. 08JDG037), and the Natural Science Foundation of Jiangsu Higher Education Institutions (Grant No. 09KJB150001). ’ REFERENCES (1) Feringa, B. L. Molecular Switches; Wiley-VCH Verlag GmbH: Weinheim, Germany, 2001. (2) (a) He, J.; Chen, F.; Liddell, P. A.; Andr
easson, J.; Straight, S. D.; Gust, D.; Moore, T. A.; Moore, A. L.; Li, J.; Sankey, O. F.; Lindsay, S. M. Nanotechnology 2005, 16, 695. (b) Yoshida, T.; Arishima, K.; Ebisawa, F.; Hoshino, M.; Sukegawa, K.; Ishikawa, A.; Kobayashi, T.; Hanazawa, M.; Horikawa, Y. J. Photochem. Photobiol., A 1996, 95, 265. (3) (a) Feng, Y.; Yan, Y.; Wang, S.; Zhu, W.; Qian, S.; Tian, H. J. Mater. Chem. 2006, 16, 3685. (b) Miyasaka, H.; Nobuto, T.; Itaya, A.; Tamai, N.; Irie, M. Chem. Phys. Lett. 1997, 269, 281. (c) Ko, C.-C.; Lam, W. H.; Yam, V. W.-W. Chem. Commun. 2008, 5203. (4) (a) Tamai, N.; Miyasaka, H. Chem. Rev. 2000, 100, 1875and references therein. (b) Kawata, S.; Kawata, Y. Chem. Rev. 2000, 100, 1777 and references therein. (5) (a) Dietz, F.; Tyutyulkov, N. Chem. Phys. 2001, 265, 165. (b) Kawai, T.; Nakashima, Y.; Irie, M. Adv. Mater. 2005, 17, 309. (c) Takami, S.; Kawai, T.; Irie, M. Eur. J. Org. Chem. 2002, 3796. (6) (a) Kobatake, S.; Irie, M. Tetrahedron 2003, 59, 8359. (b) Dietz, F.; Tyutyulkov, N. Phys. Chem. Chem. Phys. 2001, 3, 4600. (c) Tamai, N.; Saika, T.; Shimidzu, T.; Irie, M. J. Phys. Chem. 1996, 100, 4689. (d) Fukaminato, T.; Irie, M. Adv. Mater. 2006, 18, 3225. (7) (a) Kondo, M.; Tada, T.; Yoshizawa, K. Chem. Phys. Lett. 2005, 412, 55. (b) Yamada, T.; Kobatake, S.; Irie, M. Bull. Chem. Soc. Jpn. 2000, 73, 2179. (c) Tsujioka, T.; Lefuji, N.; Jiapaer, A.; Irie, M.; Nakamura, S. Appl. Phys. Lett. 2006, 85, 3128. (d) Fukaminato, T.; Sasaki, T.; Kawai, T.; Tamai, N.; Irie, M. J. Am. Chem. Soc. 2004, 126, 14843. (8) (a) Lee, J.; Kwon, T.; Kim, E. Tetrahedron Lett. 2007, 48, 249. (b) Okuyama, T.; Tani, Y.; Miyake, K.; Yokoyama, Y. J. Org. Chem. 2007, 72, 1634. (c) Areephong, J.; Browne, W. R.; Katsonis, N.; Feringa, B. L. Chem. Commun. 2006, 3930. (9) (a) Morimoto, M.; Kobatake, S.; Irie, M. J. Am. Chem. Soc. 2003, 125, 11080. (b) van der Molen, S. J.; van der Vegte, H.; Kudernac, T.; Amin, I.; Feringa, B. L.; van Wees, B. J. Nanotechnology 2006, 17, 310. (10) (a) Matsuda, K.; Irie, M. J. Am. Chem. Soc. 2000, 122, 7195. (b) Irie, M.; Lifka, T.; Kobatake, S.; Kato, N. J. Am. Chem. Soc. 2000, 122, 4871. (c) Kobatake, S.; Yamada, T.; Uchida, K.; Kato, N.; Irie, M. J. Am. Chem. Soc. 1999, 121, 2380. (d) Takami, S.; Kuroki, L.; Irie, M. J. Am. Chem. Soc. 2007, 129, 7319. (e) Kobatake, S.; Yamada, M.; Yamada, T.; Irie, M. J. Am. Chem. Soc. 1999, 121, 8450. (f) Kobatake, S.; Shibata, K.; Uchida, K.; Irie, M. J. Am. Chem. Soc. 2000, 122, 12135.

’ ASSOCIATED CONTENT

bS

Supporting Information. Comparison of the geometrical parameters between theory and experiment for anti conformations of the 1,2-bis(2-methyl-5-phenyl-3-thienyl)perfluorocyclopentene molecule in the ground state (Table S1); optimized geometries of para and twist conformations of the 4o molecule in the S0, S1, and S2 states in the gas phase and in solutions (Table S2); TD-DFT/PCM singlet vertical transition energies of ground and low-lying excited states corresponding to the maximum absorption band and oscillator strengths for the open-ring isomers of 4o in the gas phase and in solutions (Table S3); 922

dx.doi.org/10.1021/jp210846b |J. Phys. Chem. A 2012, 116, 913–923

The Journal of Physical Chemistry A

ARTICLE

(44) Frisch, M. J. et al. Gaussian 03, revision D. 01; Gaussian, Inc.: Wallingford, CT, 2004. (45) Bauschlicher, C. W. Chem. Phys. Lett. 1995, 246, 40. (46) Muscat, J.; Wander, A.; Harrison, N. M. Chem. Phys. Lett. 2001, 342, 397. (47) See, for example:(a) Ma, J.; Li, S.; Jiang, Y. Macromolecules 2002, 35, 1109. (b) Pogantsch, A.; Heimel, G.; Zojer, E. J. Chem. Phys. 2002, 117, 5921. (c) Fabian, J. Theor. Chem. Acc. 2001, 106, 199. (d) Sala, F. D.; Heinze, H. H.; G€ orling, A. Chem. Phys. Lett. 2001, 339, 343. (e) Zhang, G.; Ma, J.; Jiang, Y. Macromolecules 2003, 36, 2130. (f) van Faassen, M.; de Boeij, P. L. J. Chem. Phys. 2004, 121, 10707. (48) Yanai, T.; Tew, D. P.; Handy, N. C. Chem. Phys. Lett. 2004, 393, 51. (49) Frisch, M. J. et al. Gaussian 09, revision A. 02; Gaussian, Inc.: Wallingford, CT, 2009. (50) Marder, S. R.; Perry, J. W.; Tiemann, B. G.; Gorman, C. B.; Gilmour, S.; Biddle, S.; Bourhill, G. J. Am. Chem. Soc. 1993, 115, 2524. (51) Sun, H. Macromolecules 1995, 28, 701. (52) Sun, H.; Mumby, S. J.; Maple, J. R.; Hagler, A. T. J. Phys. Chem. 1995, 99, 5873. (53) Hwang, M. J.; Stochfisch, T. P.; Hagler, A. T. J. Am. Chem. Soc. 1994, 116, 2515. (54) Meng, S.; Ma, J.; Jiang, Y. J. Phys. Chem. B 2007, 111, 4128. (55) Meng, S.; Ma, J. J. Phys. Chem. B 2008, 112, 4313. (56) Liu, Z.; Ma, J. J. Phys. Chem. A 2011, 115, 10136. (57) Zhang, G.; Pei, Y.; Ma, J.; Yin, K.; Chen, C.-L. J. Phys. Chem. B 2004, 108, 6988. (58) Zhang, G.; Ma, J. J. Phys. Chem. B 2007, 111, 11670. (59) Wen, J.; Ma, J. Theor. Comput. Chem. 2009, 8, 677. (60) Wen, J.; Ma, J. Langumuir 2010, 26, 5595. (61) Nos
e, S. Mol. Phys. 1984, 52, 255. (62) Allen, M. P.; Tildesley, D. J. Computational Simulation of Liquids; Oxford: New York, 1987. (63) Karasawa, N.; W. A. G., III. J. Phys. Chem. 1989, 93, 7320. (64) Materials studio, version 4.0; Accelrys Inc.: San Diego, CA, 2006. (65) Harris, D. C.; Bertolucci, M. D. Symmetry and Spectroscopy: An Introduction to Vibrational and Electronic Spectroscopy; Oxford University Press: New York, 1978. (66) See, for example:(a) Svishchev, I. M.; Kusalik, P. G. J. Chem. Phys. 1993, 99, 3049. (b) Gubskaya, A. V.; Kusalik, P. G. J. Phys. Chem. A 2004, 108, 7151. (c) Gubskaya, A. V.; Kusalik, P. G. J. Phys. Chem. A 2004, 108, 7165. (d) Liu, X.; Zhang, S.; Zhou, G.; Wu, G.; Yuan, X.; Yao, X. J. Phys. Chem. B 2006, 110, 12062. (e) Liu, Z.; Huang, S.; Wang, W. J. Phys. Chem. B 2004, 108, 12978. (f) Soper, A. K. J. Chem. Phys. 1994, 101, 6888. (g) Yokogawa, D.; Sato, H.; Sakaki, S. J. Chem. Phys. 2006, 125, 114102–1. (h) Vishnyakov, A.; Lyubartsev, A. P.; Laaksonen, A. J. Phys. Chem. A 2001, 105, 1702. (67) Laaksonen, L. J. Mol. Graphics 1992, 10, 33.

(11) Katsonis, N.; Kudernac, T.; Walko, M.; Molen, S. J. v. d.; Wees, B. J. v.; Feringa, B. L. Adv. Mater. 2006, 18, 1397. (12) (a) Yamada, T.; Kobatake, S.; Muto, K.; Irie, M. J. Am. Chem. Soc. 2000, 122, 1589. (b) Morimoto, M.; Kobatake, S.; Irie, M. Chem.— Eur. J. 2003, 9, 621. (13) Uchida, K.; Ishikawa, T.; Takeshita, M.; Irie, M. Tetrahedron 1998, 54, 6627. (14) Shibata, K.; Muto, K.; Kobatake, S.; Irie, M. J. Phys. Chem. A 2002, 106, 209. (15) (a) Irie, M. Chem. Rev. 2000, 100, 1685. and references therein. (b) Irie, M.; Uchida, K. Bull. Chem. Soc. Jpn. 1998, 71, 985. (c) Irie, M.; Lifka, T.; Uchida, K.; Kobatake, S.; Shindo, Y. Chem. Commun. 1999, 747. (16) Takeshita, M.; Kato, N.; Kawauchi, S.; Imase, T.; Watanabe, J.; Irie, M. J. Org. Chem. 1998, 63, 9306. (17) Irie, M.; Miyatake, O.; Uchida, K.; Eriguchi, T. J. Am. Chem. Soc. 1994, 116, 9894. (18) Li, J.; Speyer, G.; Sankey, O. F. Phys. Rev. Lett. 2004, 93, 248302. (19) Guillaumont, D.; Kobayashi, T.; Kanda, K.; Miyasaka, H.; Uchida, K.; Kobatake, S.; Shibata, K.; Nakamura, S.; Irie, M. J. Phys. Chem. A 2002, 106, 7222. (20) Duli
c, D.; van der Molen, S. J.; Kudernac, T.; Jonkman, H. T.; de Jong, J. J. D.; Bowden, T. N.; van Esch, J.; Feringa, B. L.; van Wees, B. J. Phys. Rev. Lett. 2003, 91, 207402. (21) Ern, J.; Bens, A. T.; Martin, H.-D.; Mukamel, S.; Schmid, D.; Tretiak, S.; Tsiper, E.; Kryschi, C. Chem. Phys. 1999, 246, 115. (22) Hania, P. R.; Telesca, R.; Lucas, L. N.; Pugzlys, A.; van Esch, J.; Feringa, B. L.; Snijders, J. G.; Duppen, K. J. Phys. Chem. A 2002, 106, 8498. (23) Nakamura, S.; Irie, M. J. Org. Chem. 1988, 53, 6136. (24) (a) Jacquemin, D.; Perpete, E. A.; Maurel, F.; Perrier, A. J. Phys. Chem. C 2010, 114, 9489. (b) Perrier, A.; Maurel, F.; Aubard, J. J. Photochem. Photobiol., A 2007, 189, 167. (25) (a) Feng, Y. L.; Zhang, Q.; Tan, W. J.; Zhang, D. Q.; Tu, Y. Q.; Ågren, H.; Tian, H. Chem. Phys. Lett. 2008, 455, 256. (b) Laurent, A. D.; Andr
e, J.-M.; Perpete, E. A.; Jacquemin, D. J. Photochem. Photobiol., A 2007, 192, 211. (26) Irie, M.; Sayo, K. J. Phys. Chem. 1992, 96, 7671. (27) Corredor, C. C.; Belfield, K. D.; Bondar, M. V.; Przhonska, O. V.; Hernandez, F. E.; Kachkovsky, O. D. J. Photochem. Photobiol., A 2006, 184, 177. (28) Kim, E.; Choi, K. H.; Rhee, S. B. Macromolecules 1998, 31, 5726. (29) Chen, B.; Wang, M.; Li, C.; Xia, H.; Tian, H. Synth. Met. 2003, 135
136, 491. (30) Chen, D.-Z.; Wang, Z.; Zhao, X.; Hao, Z. J. Mol. Struct.: THEOCHEM 2006, 774, 77. (31) Kasatani, K.; Kambe, S.; Irie, M. J. Photochem. Photobiol., A 1999, 122, 11. (32) Irie, M.; Mohri, M. J. Org. Chem. 1988, 53, 803. (33) Miyasaka, H.; Nobuto, T.; Murakami, M.; Ttaya, A.; Iamai, N.; Irie, M. J. Phys. Chem. A 2002, 106, 8096. (34) Uchida, K.; Nakayama, Y.; Irie, M. Bull. Chem. Soc. Jpn. 1990, 63, 1311. (35) Tomasi, J.; Persico, M. Chem. Rev. 1994, 94, 2027 and references therein. (36) Cossi, M.; Barone, V. J. Chem. Phys. 2000, 112, 2427. (37) Cossi, M.; Barone, V.; Cammi, R.; Tomasi, J. Chem. Phys. Lett. 1996, 255, 327. (38) Mennucci, B.; Cances, E.; Tomasi, J. J. Phys. Chem. B 1997, 101, 10506. (39) Cances, E.; Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997, 107, 3032. (40) Cossi, M.; Barone, V. J. Chem. Phys. 1998, 109, 6246. (41) Mennucci, B.; Cammi, R.; Tomasi, J. J. Chem. Phys. 1999, 110, 6858. (42) Cammi, R.; Mennucci, B. J. Chem. Phys. 1999, 110, 9877. (43) Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. J. Chem. Phys. 2001, 114, 5691. 923

dx.doi.org/10.1021/jp210846b |J. Phys. Chem. A 2012, 116, 913–923