Observation of Evidence for the ÏâÏ Hyperconjugation in the S1

Article pubs.acs.org/JPCA

Observation of Evidence for the π*−σ* Hyperconjugation in the S1 State of o-, m-, and p-Fluorotoluenes by Double-Resonance Infrared Spectroscopy Takashi Chiba,† Katsuhiko Okuyama,‡ and Asuka Fujii*,† †

Department Chemistry, Graduate School of Science, Tohoku University, Sendai 980-8578, Japan Department of Chemical Biology and Applied Chemistry, College of Engineering, Nihon University, Koriyama 963-8642, Japan

‡

S Supporting Information *

ABSTRACT: Drastic changes of the methyl internal rotation potential energy functions upon the electronic excitation have been reported for o- and m-fluorotolunes [Okuyama, K.; Mikami, N.; Ito, M. J. Phys. Chem. 1985, 89, 5617−5625], and their physical origin has been attributed to the π*−σ* hyperconjugation. To observe direct evidence of the π*−σ* hyperconjugation, double-resonance infrared spectroscopy was carried out in the CH stretching vibrational region in both the S0 and S1 states of jet-cooled o-, m-, and p-fluorotoluenes. In the spectra of both o- and m-fluorotoluenes, some of the methyl CH bands were red-shifted upon the electronic excitation while the residual CH bands stayed in the same frequency region. The normal-mode analysis demonstrated that the shift behavior correlates to the relative conformation between the methyl CH bond and the phenyl ring plane. This conformation-dependent methyl CH bond weakening clearly supports the presence of the π*−σ* hyperconjugation in o- and m-fluorotoluenes. The similar red-shift of the methyl CH bands upon the electronic excitation was seen also in p-fluorotoluene though the magnitude of the shift was much smaller. The mechanism of its internal rotation potential energy behavior, however, can be different from those of the o- and m-isomers.

I. INTRODUCTION Internal rotation of the methyl group in substituted toluenes is one of prototypes of large amplitude motions in polyatomic molecules. Drastic changes of its potential energy function upon the electronic excitation have been of great interest in both experimental and theoretical studies. This problem has been first reported by Okuyama et al. in 1985.1 They have measured the fluorescence excitation and single vibronic level dispersed fluorescence spectra of the S1−S0 transition of jetcooled o-, m-, and p-fluorotoluenes (FTs) and have determined the internal rotation potential energy functions of the methyl group in the S0 and S1 states. As a result, they have found the significant changes of the potential energy functions upon the electronic excitation, and these changes seem to be strongly against the chemical intuition. The behavior of the methyl internal rotation potential energy functions of FTs is summarized in Figure 1. In o-FT, the potential energy barrier in the S0 state is 228 cm−1, which would be reasonably interpreted by the steric repulsion due to the neighboring fluorine atom. The potential energy barrier, however, drastically decreases to only 22 cm−1 in the S1 state, and the methyl group is regarded as an almost free rotor in spite of the fluorine atom at the o-position. In contrast, the potential energy barrier of the m-isomer is 16 cm−1 in the S0 state, reflecting the large distance between the methyl group and the fluorine atom. The potential energy barrier, however, remarkably increases up to 124 cm−1 with the electronic excitation, and this makes the methyl group a hindered rotor in the S1 state. In addition, the p-isomer also © XXXX American Chemical Society

shows the change of the potential energy barrier upon the electronic excitation though the magnitude of the change is much smaller (from 5 to 34 cm−1) than those in the other two isomers. Since this first report, the similar potential energy behavior of the methyl rotation upon the electronic excitation has been observed in many substituted toluenes2−10 and related molecules.11−16 The physical origin of these drastic changes upon the electronic excitation has been studied theoretically by Nakai and co-workers.17−20 They have calculated the internal rotation potential energy functions of o- and m-FTs in the S0 and S1 states with the HF/CIS levels, and have succeeded in quantitatively reproducing the observed functions.17,18 They have also calculated orbital energy behavior along the rotational angle of the methyl group, and they have concluded that the stabilization by the hyperconjugation between the aromatic π* orbital and the methyl CH σ* orbital in LUMO plays the major role in the potential energy barrier change in both o- and mFTs. They have called this interaction π*−σ* hyperconjugation. Here we should note that this hyperconjugation occurs only in the S1 state, in which the π* orbitals are occupied by the (π, π*) transition. Though contribution of hyperconjyugation (orbital interactions) to methyl internal rotation has been pointed out for the electronic ground state at an earlier stage,21 Received: May 23, 2016 Revised: June 29, 2016

A

DOI: 10.1021/acs.jpca.6b05171 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Figure 1. Methyl internal rotation potential functions of o-, m-, and p-fluorotoluenes (FTs) in the (upper) S1 and (lower) S0 states: (a, b) o-FT, (c, d) m-FT, and (e, f) p-FT. In the plots, we define θ = 0 as one methyl C−H bond is in the phenyl ring plane (oriented to the fluorine atom). The data are taken from ref 1.

the concept of the π*−σ* hyperconjugation is unique for the electronic excited state. Nakai and co-workers have suggested that this hyperconjugation occurs also in p-FT, but the discussion on p-FT has been limited to the qualitative level.18 The theoretical analysis by Nakai et al. explains well the observed rotational potential energy behavior of substituted toluenes. However, the contribution of LUMO is a part of the total electronic energy in the S1 state. Therefore, the observed rotational potential energy function, which corresponds to the dependence of the total electronic energy on the methyl rotation under the adiabatic approximation, does not directly demonstrate the π*−σ* hyperconjugation. It has been therefore requested to provide alternative experimental evidence of the hyperconjugation in the electronic excited state. In the present study, to test the interpretation by Nakai et al., we measure infrared (IR) spectra of CH stretching vibrations of the methyl group of jet-cooled o-, m-, and p-FTs in the S0 and S1 states. In both electronic states, the vibrational ground (0a1) level is carefully selected to be observed by using the infrared− ultraviolet (IR−UV) or UV−IR double-resonance spectroscopic techniques.22,23 In the S1 ground vibrational level, the conformation of the methyl group is practically restricted to the most stable one in which Nakai et al. predicted the strong π*−σ* hyperconjugation in o- and m-FTs.17,18 If the π*−σ* hyperconjugation occurs in the S1 state, it will reduce strength of methyl CH bonds by the partial electron transfer from the π* orbital of the aromatic ring to the antibonding (σ*) orbital of the CH bond. Therefore, red-shifts of methyl CH stretching bands upon the electronic excitation are expected as experimental evidence of the π*−σ* hyperconjugation in oand m-FTs. Also for p-FT, we calculate the energy of LUMO and the total potential energy function and predict the conformation of the methyl group which has the π*−σ* hyperconjugation. We interpret observed IR spectra of p-FT with these calculations. It is also interesting to observe influence of changes in the methyl orientation on the π*−σ* hyperconjugation. When the methyl group rotates in the S1 state, the methyl CH bonds experiences also conformations in which the π*−σ* hyperconjugation is spatially prohibited. This will reduce the rotationally averaged magnitude of the hyperconjugation and result in the reduction of the CH frequency shifts upon the electronic excitation. Such a situation should occur when the

internal rotation of the methyl group is excited to energy levels that are over the barrier of the rotational potential energy curve. Therefore, we also measure IR spectra at the overtone levels of 2e and 3a1 in the S1 state of o-FT, which are considered to be free rotor states for the methyl group.1

II. EXPERIMENTAL SECTION Sample vapor (o-, m-, or p-FT) was seeded in the He buffer gas. The gaseous mixture was expanded into a vacuum chamber through a supersonic jet nozzle (General Valve, series 9). The stagnation pressure of the gas was ∼3 atm. The jet expansion was skimmed to form a molecular beam. The sample of o-FT was purchased from the Kanto Kagaku Co., and those of m- and p- FTs were purchased from the Tokyo Kasei Co. These samples were used without further purification since their electronic spectra have been reported.1 The first UV laser pulse was introduced into the interaction region from the perpendicular direction to the molecular beam, and it excited the molecules to the S1 state. The second UV laser pulse, which was spatially overlapped with the first one, successively ionized the excited molecules. The wavelength of the second UV laser was set to 285 nm. The produced ions were mass-selected by the time-of-flight type mass spectrometer, and the ion intensity was monitored. The UV lasers were the second harmonics of the outputs of pulsed dye lasers (Laser Analytical Systems, LDL20505) pumped by Nd:YAG lasers (Continuum, Surelite III), which were synchronized by a pulsed delay generator (Stanford Research Systems, DG535). Typical power was ∼0.01 and ∼0.05 mJ/pulse for the first and second UV lasers, respectively. Nonliner optical crystals (β-BBO) were used to generate the second harmonics. IR Spectroscopy in S0. We employed IR−UV doubleresonance spectroscopy to observe IR spectra of the jet-cooled samples at the ground vibrational level (0a1) in the S0 state.22 A tunable IR laser pulse was introduced 50 ns prior to the first UV laser pulse of which wavelength was fixed to excite to the origin band of the S1−S0 transition. The origin band wavelength of each sample is given in ref 1. The second UV laser pulse was introduced without the delay time to the first UV laser pulse. In this situation, the IR light absorption leads to depletion of the produced ion intensity due to the reduction of the vibrational ground level population. By monitoring the ion B


Article


IV. RESULTS AND DISCUSSION o- and m-Fluorotoluenes. The observed IR spectra of oand m-FTs in the S0 and S1 states are shown in Figure 2 with

intensity while scanning the IR frequency, an IR spectrum of the S0 vibrational ground level was measured as an ion dip spectrum. IR Spectroscopy in S1. For measurements of IR spectra in the S1 state, we employed the scheme of UV−IR doubleresonance spectroscopy.22,23 In this scheme, a tunable IR laser pulse was introduced 10 ns after the irradiation of the first UV laser pulse, of which the wavelength was carefully tuned at the S1−S0 origin band. The second UV laser pulse was introduced 10 ns after the IR laser pulse. In this situation, the IR light absorption in the S1 state leads to reduction of the produced ion intensity because the vibrational excitation accelerates the nonradiative decay process in the S1 state. An IR spectrum of the S1 vibrational ground level was measured as an ion dip spectrum. We excited o-FT molecules also to the overtone levels of 2e or 3a1 in the S1 state by tuning the first UV laser wavelength1 and applied the same scheme of UV-IR spectroscopy to observe IR spectra of these methyl rotation levels. The IR light was generated by difference frequency mixing between the fundamental output of a dye laser (Continuum ND 6000) and the second harmonics of a Nd:YAG laser (Continuum PL-8000). A nonlinear optical crystal (LiNbO3) was used for the mixing. Typical IR output power was ∼0.3 mJ/ pulse. All the observed IR spectra were normalized by the IR power.

III. COMPUTATIONS Quantum chemical calculations were performed with the Gaussian 09 program suite.24 For the S0 and S1 states, we used the HF/6-31G(d,p) and CIS/6-31G(d,p) levels of theory, respectively. By these levels, we calculated energy-optimized structures, energy of LUMO and total energy potential curves along the methyl rotation, and vibrational spectra, which are directly compared with the observed ones. The calculated vibrational spectra were obtained by the normal-mode analysis. We applied a scaling factor of 0.9153 to all the calculated harmonic spectra. Visualization of normal modes was performed by WebMO.25 The levels of theory employed in this study would be rather primitive in the modern standard of quantum chemical calculations. We, however, dared to employ these levels in this study with the following reasons. Though CIS is one of the most basic levels of theory for the S1 state, Nakai et al., in their original study in 1999,17 have quantitatively reproduced the observed methyl rotation potential energy behavior upon the electronic excitation by the CIS/6-31G(d,p) level calculations, and they have successfully interpreted the potential energy behavior in terms of the π*−σ* hyperconjugation. This suggests that the CIS level is essentially enough to analyze the π*−σ* hyperconjugation. To test the interpretation by Nakai et al., the same level of theory should be chosen to simulate IR spectra. Frequency shifts of methyl CH stretching bands should be reproducible by the CIS level if the shifts are induced by the π*−σ* hyperconjugation. In other words, we cannot exclude effects other than the π*−σ* hyperconjugation even if we can reproduce observed methyl CH shifts with use of levels of theory beyond CIS. Therefore, the CIS/6-31G(d,p) level was chosen for the simulation in the S1 state in this study. Being consistent with this choice for the S1 state, we chose the HF/6-31G(d,p) level for calculations of the S0 state because use of this level for the S0 state corresponds to use of CIS in the S1 state.

Figure 2. Observed and calculated (stick) IR spectra of o-FT in (a) S0 and (b) S1 and m- FT in (c) S0 and (d) S1.

their calculated stick spectra. In the measurements of these spectra, the first UV laser frequency was carefully tuned to the origin band of the electronic transition,1 and the IR spectra of the vibrationally ground levels were observed in both the S1 and S0 states. Here we should note that the methyl rotational angle is localized at around the potential minimum position in the vibrationally ground levels. Based on typical vibrational frequencies of functional groups, the bands higher than ∼3020 cm−1 are assigned to CH stretching vibrations of the phenyl ring moiety and those lower than 3020 cm−1 are attributed to CH stretches of the methyl group. In both the phenyl and methyl CH regions, the number of the observed bands is larger than the number of the CH oscillators in the functional group. The CH stretch regions in the observed spectra suffer anharmonic coupling with some overtone/combination bands, and this prevents us from simple assignments of the observed bands. It is however, reasonable to assume that the deperturbed (zero-order) band frequency locates at around the observed intense band, and we can qualitatively assign the CH stretch bands of the methyl groups. Assignments of the phenyl CH stretches are more difficult. In this article, however, the phenyl CH stretches are out of focus because they would not strongly participate in the hyperconjugation. In the observed spectrum of the S0 state of o-FT, the bands in the 2920−3000 cm−1 are clearly attributed to the methyl CH C


Article

The Journal of Physical Chemistry A stretches. On the other hand, the spectrum of its S1 state shows strong bands in 2820−2920 cm−1 as well as a band at ∼3000 cm−1. Both of bands should be assigned to the methyl CH stretches. This spectral change demonstrates that the major part of the methyl CH stretch bands shows clear red-shifts upon the electronic excitation to the S1 state while one CH band remains in almost the same frequency region. This suggests that there are two types of the methyl CHs; one is largely influenced by the electronic excitation of the phenyl ring moiety, and the other is not. The same shift trends are seen in the spectra of m-FT, and the trends are even clearer. No strong absorption is seen in the spectrum below 2900 cm−1 in the S0 state, but some remarkable bands appear in this region upon the electronic excitation. In addition, one band still remains at around 3000 cm−1 in the S1 spectrum. Here we should note that the magnitude of the red-shift is also similar between o-and m-FTs. The observed band frequencies and tentative assignments are summarized in Tables S1 and S2 in the Supporting Information. The simulated IR spectra of the S0 and S1 states of o- and mFTs are also shown in Figure 2. The levels of theory of the simulations are HF/6-31G(d,p) and CIS/6-31G(d,p) for the S0 and S1 states, respectively. As stated in the Computations section, these levels of theory have been employed by Nakai and co-workers to quantitatively reproduce the experimentally determined internal rotation potential energy functions of the methyl group.17,18 This means that the physical origin of the potential energy behavior upon the electronic excitation should be well involved in these levels of theory. We also confirmed that our calculations on the internal rotation potential energy functions of o- and m-FTs reproduce well the results reported by Nakai et al.17,18 We actually found some very small differences of the calculated potential energy barrier height between our results and those by Nakai et al. (a few cm−1 in the potential function of the S0 state and ∼10 cm−1 in the S1 state) in spite of the same levels of theory. The improved optimization precision of the Gaussian program and removal of the frozen core approximation employed by Nakai et al. in the CIS calculations might contribute to the minor differences in these calculation results. We used the most stable structures (methyl conformation) in the S0 and S1 states at these levels for the harmonic vibrational simulations. Because of the limitation of the harmonic approximation (lack of anharmonic coupling), the agreement between the observed and simulated spectra is rather qualitative. However, the simulated spectra support well the band assignments given above. The assignment of the band at around 3000 cm−1 to the methyl CH is especially evidenced by the simulations. Moreover, the simulated spectra clearly reproduce the observed red-shift trends of the methyl CH stretches upon the electronic transition and the two types of the shift behavior. Figure 3 shows the calculated normal modes of the methyl CH stretches of o- and m-FTs. We note that these modes are based on the methyl conformation at the potential energy minimum. Because of the approximate C3 symmetry of the three CH bonds, the CH stretch bands are composed of one symmetric (sym) and two antisymmetric (antisym 1 and 2) modes. In Figure 2, the estimated correspondence between the simulated and observed bands is indicated by the dashed lines. As seen in the figure, in both o- and m-FTs, the sym and antisym 1 bands shift to red upon the electronic excitation but the antisym 2 band does not. Nakai et al. have demonstrated

Figure 3. Normal modes of CH stretching vibrations of o- and m-FTs in the S0 and S1 states.

that in both o- and m-FTs the hyperconjugation occurs among σ* orbitals of out-of-plane methyl C−H bonds and π* orbitals of the phenyl ring in the S1 state, while orbitals of in-plane methyl C−H bonds do not participate in this interaction.17,18 This spatial selectivity of the interaction suggests that only vibrational modes that include out-of-plane CH stretches show red-shifts upon the electronic excitation. As seen in Figure 3, the sym and antisym 1 modes are out-of-plane CH stretch while the antisym 2 mode is in-plane CH stretch. Therefore, the observed and calculated shift trends of the methyl CH stretch bands upon the electronic excitation are well explained by the π*−σ* hyperconjugation in both o- and m-FTs. Here, we discuss about the correlation between the amount of the CH band shift and the magnitude of the π*−σ* hyperconjugation. It would be reasonable to suppose that the amount of the band shift increases with increasing magnitude of the interaction. o- and m-FTs show the quite similar amounts of the CH band shifts upon the electronic excitation, and the observed shifts are ca. 70−80 cm−1 in both the sym and antisym 1 CH bands. The simulated frequencies also support well this observation. On the other hand, to define the magnitude of the π*−σ* hyperconjugation, the LUMO energy behavior about the methyl internal rotation is a good measure because this interaction occurs in LUMO. When the hyperconjugation occurs, the LUMO energy is lowered while the energy is relatively raised when the hyperconjugation does not occur. Therefore, we define the magnitude of the π*−σ* hyperconjugation as the difference between the maximum and minimum LUMO energies about the methyl internal rotation. In this definition, from the calculations of Nakai and coworkers, the magnitude of the hyperconjugation in o-FT is evaluated to be about 1100 cm−1 and that of m-FT is about 1200 cm−1.17,18 Therefore, both isomers have almost the same magnitude of the interaction. The similar amounts of the observed band shifts of o- and m-FTs are consistent with the magnitude of the π*−σ* hyperconjugation evaluated by by Nakai and co-workers. We should note that the magnitude of the π*−σ* hyperconjugation would directly correlate to the frequency of the local out-of-plane CH modes. The normal modes are combinations of the local modes and their frequency shifts do not directly correspond to those of the local CH modes. However, the basic skeletal structure of the methyl group is same in o- and m-FTs, and this is kept even after the electronic D


Article


comparison with the spectrum of the 0a1 level, there are two changes in the spectra of the 2e and 3a1 levels. First, the antisym 2 band does not appear at ca. 3000 cm−1 in the 2e and 3a1 spectra. Instead, bands newly appear at around 2985 cm−1. These new bands can be assigned to antisym 2 (and its anharmonic coupling bands) at the 2e and 3a1 levels. The antisym 2 band is red-shifted upon the internal rotational excitation. Second, the bands at around 2880 cm−1, which correspond to the sym and antisym 1 bands, change their shape. The intensity of the high frequency side becomes larger in the 2e and 3a1 spectra. This change can be interpreted as the small blue-shifts of the sym and antisym 1 bands. These changes indicate the strong coupling between the internal rotation and CH stretches of the methyl group. When the conformation of the methyl group changes along its internal rotation, the potential energy of the methyl CH stretching modes also changes. The mechanism of the coupling can be explained by the π*−σ* hyperconjugation. Along with the internal rotation, an out-of-plane CH bond under the hyperconjugation becomes an in-plane CH bond which is free from the hyperconjugation, and vice versa. The methyl rotation averages the out-of-plane and in-plane CH stretches, and their frequencies become closer to each other. This mechanism explains the opposite shift trends of the sym/ antisym 1 bands and the antisym 2 band upon the internal rotation excitation. Therefore, these results also support the existence of the π*−σ* hyperconjugation. p-Fluorotoluene. In contrast with the cases of o- and mFTs, the influence of the electronic excitation to the methyl rotation potential energy function is very weak in p-FT. Okuyama et al. have observed that the methyl rotation of p-FT is practically free in the S0 state, and a very small barrier occurs in the S1 state.1 This methyl rotational potential behavior strongly suggests that the importance of the π*−σ* hyperconjugation is much less in p-FT. The observed IR spectra of p-FT in the S0 and S1 states are shown in Figure 6 with their calculated stick spectra. In the observed spectra of p-FT, the number of observed bands is obviously greater than that of CH oscillators in the molecule. This means that anharmonic coupling among CH stretch fundamentals and overtones/combinations of other modes extensively occurs. Assignment of each band is quite difficult at

excitation. The construction of the normal modes based on the local CH modes is almost common in o- and m-FTs. Therefore, when the magnitude of the hyperconjugation is similar between o- and m-FTs in the S1 state, the amount of the band shift is also similar between them. The observed spectra suffer the strong anharmonic coupling, and this prevents from definite assignments of the methyl CH bands and the evaluation of the precise frequency shifts upon the electronic excitation. The results of the above discussion are only semiquantitative. However, the observed and calculated spectral behavior of the methyl CH stretches is well explained by the π*−σ* hyperconjugation theory of Nakai and coworkers.17,18 The present results strongly support the existence of the π*−σ* yperconjugation in o- and m-FTs. Influence of the Methyl Internal Rotation Excitation. As described in the above subsection, the π*−σ* hyperconjugation influences only on out-of-plane CH bonds of the methyl group. In the ground vibrational (0a1) level of the S1 state of o-FT, as seen in the potential energy function of the internal rotation in Figure 4, the methyl conformation relative

Figure 4. Energy levels about internal rotation in the S1 state of o-FT. The energies of the overtone levels of 2e and 3a1 are higher than the barrier of the rotational potential energy. The data are taken from ref 1.

to the phenyl ring plane is almost fixed, and the out-of-plane and in-plane CH bonds can be distinguished. If the internal rotation of the methyl group is excited, the methyl conformation is no longer fixed and the influence of the hyperconjugation upon the methyl group should be changed. To examine the effect of the methyl internal rotation excitation, we excited the overtone levels of 2e and 3a1 in the S1 state which can be regarded as free rotor levels (see Figure 4), and we observed IR spectra at these levels. The S1−S0 transition frequencies to these levels are seen in ref 1. Figure 5 shows the comparison of the observed infrared spectra at the 0a1, 2e, and 3a1 levels in the S1 state of o-FT. In

Figure 6. Observed and calculated IR spectra of p-FT in (a) S0 and (b) S1.

Figure 5. IR spectra of (a) 0a1, (b) 2e, and (c) 3a1 levels in the S1 state of o-FT. Spectrum (a) is reproduction of that in Figure 2b. E


Article


The present calculation results qualitatively reproduce the observed trends in the methyl rotation potential energy functions and CH stretch vibrations. This means that the physical essence of the methyl rotation behavior upon the electronic excitation in p-FT might be involved in the calculations at the HF/CIS level as well as the cases of oand m-FTs. The energy curve of LUMO along the methyl rotational angle is shown in Figure 7c. In this plot, the minimum energy is found at θ = 0° and the maximum energy is at θ = 30°. The orbital shapes of LUMO at these angles are shown in Figure 8. The orbital overlap between π* and σ*

the present stage, and we can only distinguish the methyl and phenyl CH stretch regions. The observed and calculated band frequencies are summarized in Table S3 of the Supporting Information. In the gross feature of the methyl CH stretch bands, a redshift trend upon the electronic excitation is seen. However, the magnitude of the shift is clearly much less than those in o- and m-FTs. This is consistent with the expectation from the methyl rotation potential energy behavior. In addition, the calculated spectra also reproduce the weak red-shift trend of the methyl CH stretch bands. Both observed and simulated spectra imply that the influence of the π*−σ* hyperconjugation in p-FT is practically much weaker than o- and m-FTs. In the previous theoretical study by Nakai et al., p-FT was out of focus, and only its small methyl rotational barriers in the S0 and S1 states have been shown in their calculations.18 Here, to examine the role of the π*−σ* hyperconjugation in p-FT, we calculate the electronic energy in the S0 and S1 states and the LUMO energy along the rotational angle of the methyl group. As seen in Figures 7a and 7b, the calculated methyl rotation

Figure 8. LUMO orbitals in p-FT at (a) θ = 0° and (b) θ = 30°.

occurs at θ = 0°, the energy minimum angle, and the overlap is very small at θ = 30°, the energy maximum. Therefore, also in p-FT, the π*−σ* hyperconjugation occurs, and it determines the methyl rotational angle dependence of the LUMO energy as well as the cases of o- and m-FTs. This is consistent with the observed red-shift trend of the methyl CH stretches though we should note that the magnitude of the shift is much smaller than o- and m-FTs. Comparing with the methyl internal rotation potential curve in the S1 state and the LUMO energy curve along the methyl internal rotation, however, we find that the minimum energy angle of the methyl rotation potential energy function in the S1 state (θ = 30°) corresponds to the maximum of the LUMO energy (i.e., minimum of the magnitude of the π*−σ* hyperconjugation). This result indicates that the internal rotation potential energy function in p-FT is actually governed by factors other than the π*−σ* hyperconjugation since the magnitude of the hyperconjugation is quite small. However, the origin of weak red-shift trend in pFT can also be interpreted as the π*−σ* hyperconjugation. The methyl CHs are loosely hindered in the valley of the internal rotation potential curve because of the low barrier height. Therefore, the methyl CHs can be in the position in which they interact with π* orbitals even at the ground vibrational (0a1) level.

Figure 7. Methyl rotational angle dependence of (a) potential energy in S0, (b) potential energy in S1, and (c) LUMO energy of p-FT.

potential energy barriers are 2 and 10 cm−1 in the S0 and S1 states, respectively, and the potential energy minimum is found at θ = 30° in both the states. In the previous calculations by Nakai et al., the barrier height was evaluated to be 5 and 2 cm−1, respectively, and the minimum was found at θ = 0° in both the states.18 The experimental potential barriers measured by Okuyama et al. are 4.8 and 33.7 cm−1 in the S0 and S1 states, respectively.1 (No rotation of the energy minimum angle upon the electronic excitation was also confirmed though the absolute angle cannot be determined by the experiment.) The present calculations show much better agreement with the observed trend in the potential energy behavior. The reason for this difference with the results of Nakai et al. might be same as those in o- and m-FTs (precision of optimization and frozen core approximation). Since the barrier height is very small in pFT, small error can largely change the potential shape.

V. CONCLUSION We measured the IR spectra of o- and m-FTs in the S0 and S1 states. The CH stretching vibrational region was observed, and the shifts of the methyl CH bands upon the electronic excitation were well interpreted by the π*−σ* hyperconjugation in the S1 state. The present observation strongly supports the π*−σ* hyperconjugation mechanism, which has been proposed to interpret the methyl rotation potential energy F


Article


(6) Zhao, Z. Q.; Parmenter, C. S.; Moss, D. B.; Bradley, A. J.; Knight, A. E.W.; Owens, K. G. p-fluorotoluene. I. Methyl (CH3 and CD3) Internal Rotation in the S1 and S0 States. J. Chem. Phys. 1992, 96, 6362−6377. (7) Takazawa, K.; Fujii, M.; Ito, M. Internal Rotation of the Methyl Group in Fluorotoluene Cations as Studied by Pulsed Field IonizationZero Kinetic Energy Spectroscopy. J. Chem. Phys. 1993, 99, 3205− 3217. (8) Fujii, M.; Yamauchi, M.; Takazawa, K.; Ito, M. Electronic Spectra of o-, m- and p-TolunitrileSubstituent Effect on Internal Rotation of the Methyl Group. Spectrochim. Acta, Part A 1994, 50, 1421−1433. (9) Borst, D. R.; Pratt, D. W. Toluene: Structure, Dynamics, and Barrier to Methyl Group Rotation in Its Electronically Excited State. A Route to IVR. J. Chem. Phys. 2000, 113, 3658−3669. (10) Alvarez-Valtierra, L.; Yi, J. T.; Pratt, D. W. Rotationally Resolved Electronic Spectra of 2- and 3-Methylanisole in the Gas Phase: A Study of Methyl Group Internal Rotation. J. Phys. Chem. B 2006, 110, 19914−19922. (11) Breen, P. J.; Warren, J. A.; Bernstein, E. R.; Seeman, J. I. A Study of Nonrigid Aromatic Molecules by Supersonic Molecular Jet Spectroscopy. I. Toluene and the Xylenes. J. Chem. Phys. 1987, 87, 1917−1926. (12) Held, A.; Selzle, H. L.; Schlag, E. W. Methyl Group Rotational Dynamics in o-, m-, and p-Xylene Cations from Pulsed Field Ionization Zero-Kinetic-Energy Spectroscopy. J. Phys. Chem. A 1998, 102, 9625− 9630. (13) Mordziń ski, A.; Leś, A.; Stepanenko, Y.; Rycombel, J.; Adamowicz, L. S0 and S1 Spectroscopy of Jet Cooled 9-Cyano-10Methylanthracene: The Methyl Group as a Molecular Rotor. J. Mol. Spectrosc. 2005, 233, 98−109. (14) Nakagaki, M.; Nishi, E.; Sakota, K.; Nishi, K.; Nakano, H.; Sekiya, H. Internal Rotation of Methyl Group in 2- and 1Methylanthracene Studied by Electronic Spectroscopy and DFT Calculations. Chem. Phys. 2005, 316, 178−184. (15) Nakagaki, M.; Nishi, E.; Sakota, K.; Nakano, H.; Sekiya, H. A Model Two-Dimensional Potential for Internal Rotation of 9Methylanthracene Studied by Electronic Spectroscopy and DFT Calculations. Chem. Phys. 2006, 328, 190−196. (16) Baba, M.; Mori, K.; Saito, M.; Kowaka, Y.; Noma, Y.; Kasahara, S.; Yamanaka, T.; Okuyama, K.; Ishimoto, T.; Nagashima, U. CH3 Internal Rotation in the S0 and S1 States of 9-Methylanthracene. J. Phys. Chem. A 2010, 113, 2366−2371. (17) Nakai, H.; Kawai, M. Nature of the Change in the Rotational Barrier of the Methyl Group Due to S0 - S1 Excitation. Chem. Phys. Lett. 1999, 307, 272−276. (18) Nakai, H.; Kawai, M. π*−σ* Hyperconjugation Mechanism on the Rotational Barrier of the Methyl Group (I): Substituted Toluenes in the Ground, Excited, and Anionic States. J. Chem. Phys. 2000, 113, 2168−2174. (19) Nakai, H.; Kawamura, Y. π*−σ* Hyperconjugation Mechanism on the Rotational Barrier of the Methyl Group (II): 1- and 2Methylnaphthalenes in the S0, S1, C0, and A1 States. Chem. Phys. Lett. 2000, 318, 298−304. (20) Kawamura, Y.; Nagasawa, T.; Nakai, H. π* − σ* Hyperconjugation Mechanism on the Rotational Barrier of the Methyl Group (III): Methyl-Azabenzenes in the Ground, Excited, and Anionic states. J. Chem. Phys. 2001, 114, 8357−8363. (21) Hehre, W. J.; Pople, J. A.; Devaquet, A. J. P. Torsional Potentials of Methyl Rotors Attached to Polar Linkages. J. Am. Chem. Soc. 1976, 98, 664−668. (22) Ebata, T.; Fujii, A.; Mikami, N. Vibrational Spectroscopy of Small-Sized Hydrogen-Bonded Clusters and Their Ions. Int. Rev. Phys. Chem. 1998, 17, 331−361. (23) Walther, Th.; Bitto, H.; Minton, T. K.; Huber, J. R. UV-IR Double-Resonance Spectroscopy of Jet-Cooled Propynal Detected by the Fluorescence Dip Method. Chem. Phys. Lett. 1994, 231, 64−69. (24) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,

behavior in substituted toluenes. The observed CH frequency shifts were reproduced by the simple HF/CIS level calculations, and this demonstrates that the physical essence of the π*−σ* hyperconjugation is involved at this level of theory. For more precise and quantitative discussion, calculations at higher levels of theory, as well as involving anharmonic couplings, are highly requested in future work. We also observed the IR spectra at the overtone levels of 2e and 3a1 in the S1 state of o-FT, which are considered to be free rotor states. With the comparison to the spectrum at the 0a1 level (vibrational ground level), the strong coupling between the internal rotation and methyl CH stretching vibration was suggested. This coupling was also reasonably explained by the π*−σ* hyperconjugation. Therefore, this result also supports the contribution of the π*−σ* hyperconjugation in o-FT. For p-FT, the observed IR spectra showed that the magnitude of the π*−σ* hyperconjugation is much weaker than o- and m-FTs. The electronic potential energy in the S0 and S1 states and the LUMO energy dependence along the methyl internal rotation were calculated. The calculation results showed that the weak π*−σ* hyperconjugation also occurs in p-FT, but it does not govern the internal methyl rotation potential energy function.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpca.6b05171. Tables of the observed band frequencies, their assignments, and calculated CH stretching vibrational frequencies of o-, m-, and p-fluorotoluenes in S0 and S1; the complete author list of ref 24 (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail [email protected], Tel +81-22-795-6572 (A.F.). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We acknowledge to Dr. Toshihiko Maeyama and Dr. Yoshiyuki Matsuda for their helpful discussions. This study was supported by the Grant-in-Aid for Scientific Research (Project No. 26288002) from JSPS.

■

REFERENCES

(1) Okuyama, K.; Mikami, N.; Ito, M. Internal Rotation of the Methyl Group in the Electronically Excited State: o-, m-, and pFluorotoluene. J. Phys. Chem. 1985, 89, 5617−5625. (2) Okuyama, K.; Mikami, N.; Ito, M. Internal Rotation of the Methyl Group in the Electronically Excited State: o- and m-Toluidine. Laser Chem. 1987, 7, 197−211. (3) Mizuno, H.; Okuyama, K.; Ebata, T.; Ito, M. Rotational Isomers of m-Cresol and Internal Rotation of the CH3 Group in S0, S1, and the Ion. J. Phys. Chem. 1987, 91, 5589−5593. (4) Tanaka, S.; Okuyama, K. Internal Rotation of Methyl Group in Electronically Excited o- and m-Ethynyltoluene: New Correlation between the Hammett Substituent Constant σm and Rotational Barrier Change. J. Chem. Phys. 2011, 134, 084311−084317. (5) Hollas, J. M.; Taday, P. F. Methyl and Vinyl Torsional Potentials in cis- and trans-3-Methylstyrene from Supersonic Jet Fluorescence Spectra. J. Chem. Soc., Faraday Trans. 1991, 87, 3585−3593. G


Article

The Journal of Physical Chemistry A B.; Petersson, G. A.; et al. Gaussian 09, Revision B.01; Gaussian, Inc.: Wallingford, CT, 2009. (25) Schmidt, J. R.; Polik, W. F. WebMO, version 16.0; WebMO LLC: Holland, MI, 2015.

H


Observation of Evidence for the Ï*âÏ* Hyperconjugation in the S1

Recommend Documents

Observation of Evidence for the ÏâÏ Hyperconjugation in the S1