Photochemistry
DOI: 10.1002/anie.201004571Ordering, Interaction, and Reactivity of the Low-Lying np* and pp*
Excited Triplet States of Acetophenone Derivatives**
Sohshi Yabumoto, Shinsuke Shigeto,* Yuan-Pern Lee, and Hiro-o Hamaguchi*
The diversity of photophysics and photochemistry of the low-lying excited triplet states of aromatic carbonyl compounds has attracted considerable interest in the field of organic photochemistry.[1] For instance, the intersystem crossing
rates,[2] phosphorescence lifetimes,[3] and photoreduction
activities[4]of these compounds show a marked dependence
on both substituents and solvents. Depending on their electronic configurations, the energy of the low-lying triplet states, namely np*, pp*, and charge transfer (CT), can be influenced by substituents and solvents, with possible alter-ations in the energy-level ordering of the states. The photo-reduction proceeds via the T1state,[5]which has approximate
quantum yields that vary between 1 for the np* T1state, 0.1
for the pp* state, and 0 for the CT state.[6] The strong
substituent and solvent dependence of the photophysics and photochemistry of aromatic carbonyl compounds has thus been discussed in terms of the energy-level ordering of the np*, pp*, and CT excited triplet states.[7–10]
It is known that the photoreduction activity of aromatic carbonyl compounds varies gradually with substituents or with solvents. In particular, the pp* T1 states show varying
reactivities that cannot be accounted for solely by energy-level ordering. There have been several arguments about this reactivity variation of the pp* T1 state. It is generally
considered that the reactivity arises from mixing of the np* character into the pp* state.[1]A mechanism that involves the
thermal excitation to a closely lying np* state has also been suggested.[10–12] These arguments are not based on direct
experimental evidence on the ordering of the excited triplet
states and therefore are not conclusive; conventional spec-troscopic techniques have not been effective in observing close-lying excited triplet states of aromatic carbonyl com-pounds.[13] Thus, it is highly important to experimentally
clarify the energy-level ordering and the electronic config-urations of the low-lying excited triplet states of aromatic carbonyl compounds. We have constructed a nanosecond time-resolved absorption spectrometer that is suitable for observing the triplet–triplet transitions in the near-infrared region as well as the vibrational transitions in the mid-infrared region.[13]We have focused on the substituent dependence of
both the triplet–triplet absorption spectra and the photo-reduction activity of a series of acetophenone derivatives.
The time-resolved near-infrared spectra of acetophenone (AP) excited at 325 nm in benzene are shown in Figure 1. Upon photoexcitation, two broad transient absorption bands arise instantaneously within the time resolution of the apparatus, and decay synchronously. One band spans from 2000 to 7000 cm1, with a peak at 3500 cm1. The other band starts from 7000 cm1and extends to the higher-wavenumber
region above 12 000 cm1. The decay of these two bands is
completely synchronous with the recovery of the ground-state
Figure 1. Time-resolved near-infrared spectra of photoexcited acetophe-none in benzene. The region around 3000 cm1is blocked by the
solvent absorption signal. [*] Dr. S. Yabumoto, Prof. S. Shigeto, Prof. Y.-P. Lee, Prof. H. Hamaguchi
Institute of Molecular Science and Department of Applied Chemistry National Chiao Tung University 1001 Ta Hsueh Rd., Hsinchu (Taiwan) Fax: (+ 886) 3-572-3764
E-mail: [email protected] Homepage: http://140.113.224.181/ Prof. H. Hamaguchi
Department of Chemistry, School of Science the University of Tokyo
7-3-1 Hongo, Bukyo-ku, Tokyo (Japan) Fax: (+ 81) 3-3818-4621
E-mail: [email protected]
Homepage: http://www.chem.s.u-tokyo.ac.jp/users/struct/ [**] This work was supported by the “Aim for the Top” University Plan of
the National Chiao Tung University and the Ministry of Education, Taiwan, and in part by the National Science Council of Taiwan (grant nos. NSC-97-2811M-009-011, NSC-97-2811M-009-024, and NSC-98-2113M-009-011-MY2).
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/anie.201004571.
depletion, as monitored by vibrational bands. By analogy with benzophenone,[14]these bands are unequivocally ascribed to
electronic transitions in the triplet manifold.
The near-infrared spectra of 4’-(trifluoromethyl)aceto-phenone (CF3-AP), AP, 4’-methylacetophenone (Me-AP),
4’-methoxyacetophenone (MeO-AP), and 2-acetonaphthone (AN) in benzene at 0–1 ms (each spectrum is obtained by averaging over 100 spectra between 0 and 1 ms) are shown in Figure 2. For all the compounds studied, transient absorption bands arise instantaneously after photoexcitation and decay synchronously, as in the case of acetophenone. The spectra can be characterized by two distinct components: the lower-wavenumber band centered around 3500 cm1and the
higher-wavenumber band located above 7000 cm1. As in the case of
AP, these transient bands are ascribed to triplet–triplet transitions. The spectrum of CF3-AP shows a strong
lower-wavenumber component with a very weak higher-wavenum-ber component, while the spectrum of AN shows only a higher-wavenumber component. The spectra of the other compounds show both the lower- and higher-wavenumber components, but the intensity ratio between the two compo-nents varies markedly. Furthermore, the higher-wavenumber component seems to have a vibrational structure with a spacing of approximately 1500 cm1. The relative intensity of the lower-wavenumber component to the higher-wavenum-ber component correlates well with the photoreduction activity, as a higher intensity of the lower-wavenumber component results in a higher photoreduction activity. AN does not show any photoreduction activity in the presence of alcohols.[15]The reported photoreduction rates k
rfor CF3-AP,
AP, and Me-AP[7, 8]are given in Table 1, together with the area
intensity ratios Ir=Ilower/Ihigher observed for the four
aceto-phenone derivatives. It is clear that the reactivity of the derivatives increases as the intensity ratio is increased.
The relative energies of the lowest np* and pp* states depend on the substituent, therefore the order of the states often alters. We have found that the triplet–triplet absorption spectra of acetophenone derivatives consist of only two components, that is, the lower- and higher-wavenumber components, the intensity ratio Irof which varies markedly
with the substituent. This variation must be related to the alteration of the energy-level ordering of the lowest np* and pp* states, as suggested by the correlation of Ir with the
photoreduction activity.
It is already known that the T1state of CF3-AP has the np*
configuration and that of Me-AP and MeO-AP has the pp* configuration.[7, 8]The np* states are highly reactive toward
photoreduction, while the pp* states are much less reac-tive.[1, 4, 7, 8]It is clear from Table 1 that CF
3-AP, which has the
strongest lower-wavenumber component, shows much higher photoreduction activity than Me-AP and MeO-AP, which have weaker lower-wavenumber components. We therefore conclude that the lower-wavenumber component arises from a triplet–triplet transition from an np* state, and that the higher-wavenumber component arises from a transition from
Table 1: Photoreduction activity and band intensity ratios of substituted acetophenone derivatives.
kr[10 6
s1
m1] Ir=Ilower/Ihigher[a]
CF3-AP 6.2 [b] 22 AP 1.2[b] 6.7 Me-AP 0.13[b] 1.9 AN 0[c] 0 [a] Area intensity ratio of the lower- and higher-wavenumber compo-nents. [b] Recorded in 2-propanol/benzene binary mixtures. Values taken from reference [7]. [c] AN does not undergo photoreduction by alcohols; see reference [15].
Figure 2. Time-resolved near-infrared spectra at 0–1 ms (upper part) and energy diagrams with transitions corresponding to the bands observed in the spectra (lower) for a) AN, b) AP (c), Me-AP (b), and MeO-AP a), and c) CF3-AP.
a pp* state. Indeed, the spectrum of CF3-AP, for which the
np* T1state is much lower in energy than the pp* state, has
only the lower-wavenumber component, while the spectrum of AN, for which the pp* T1state is much lower than the np*
state, has only the higher-wavenumber band. Thus, the initial states of the lower- and the higher-wavenumber components are identified as the np* and pp* states, respectively. The final states of these transitions are likely to have the same electronic configurations as those of the initial states because the np*$np* and pp*$pp* transitions are allowed, but the np*$pp* transition is symmetry-forbidden for planar struc-tures. This assignment is consistent with the fact that the band positions and hence the energy gaps of the initial and final states do not vary much by substitution; the electronic states that have the same electronic configurations are likely to behave in a similar fashion toward substitution. The observed synchronous decay of the two components observed for AP, Me-AP, and MeO-AP indicates that the np* and pp* states are in thermal equilibrium in these three compounds. The thermal equilibration between the low-lying np* and pp* states of the acetophenone derivatives is shown in Figure 2. For AN, the lowest pp* state of which is much lower in energy than the lowest np* state, only the pp* state is populated and therefore only the higher-wavenumber component (Tpp*
band) is observed (Figure 2 a). For CF3-AP, the lowest np*
state of which is much lower than the lowest pp* state, only the np* state is populated and therefore only the lower-wavenumber component (Tnp*band) is observed (Figure 2 c).
For the molecules where the lowest np* and pp* states are close in energy, as is the case of AP, Me-AP, and MeO-AP, both the np* and pp* states are populated and therefore both the Tnp*and Tpp*bands are observed; the intensity ratios of
these bands varies with the population ratio of the np* and pp* states (Figure 2 b).
We have shown that the low-lying np* and pp* states for AP, Me-AP, and MeO-AP are close in energy and are in thermal equilibrium, which can be directly confirmed by observing the effect of temperature on the spectra. Figure 3 shows the time-resolved infrared spectra at 0–1 ms of photo-excited MeO-AP in a,a,a-trifluorotoluene (TFMT) at 29, 55, and 80 8C. It is clear that the intensity ratio kr between the
lower- and higher-wavenumber components is temperature-dependent; the higher-wavenumber component decreases while the lower component increases in intensity as the temperature increases. This result supports the hypothesis that the two triplet states are in thermal equilibrium and that the population of the lowest pp* state decreases, while that of the lowest np* increases, as the temperature increases. The enthalpy difference between the lowest np* and pp* states can be determined from the temperature dependence of the intensity ratio, assuming a Boltzmann distribution. Figure 4 shows the plot of ln(1/Ir) versus 1/T. The enthalpy difference
DH = Hnp*Hpp* between the np* T2 and pp* T1 states is
determined from the slope by a least-squares fitting of the observed points to a straight line. DH is determined to be (2.9 0.6) kJ mol1
for MeO-AP in TFMT (Figure 4 a). The entropy difference DS = Snp*Spp* determined from the
intercept is (1 5) J K1mol1; this small value is consistent
with the fact that the molecular structures are nearly the same
for the np* and pp* triplet states so that the vibrational, rotational, and translational partition functions and hence the entropy values are nearly the same for the two triplet states. The temperature dependence of the intensity ratio Irhas
also been examined for Me-AP (Figure 4 b) and AP (Fig-ure 4 c) in TFMT. The enthalpy difference is determined as
Figure 3. Time-resolved near-infrared spectra at 0–1 ms of photoexcited MeO-AP in a,a,a-trifluorotoluene at 29, 55, and 80 8C (order indicated by the arrows). The spectra are normalized by making the areas of the depletion (negative peak) of the ground-state CO stretching band (not shown) equal.
Figure 4. Relative intensities of a) MeO-AP, b) Me-AP, and c) AP plotted against 1/T.
(1.2 1.2) kJ mol1 for Me-AP. Although no reliable value
was obtained for AP because of large experimental uncer-tainties, the enthalpy difference is smaller for AP than for Me-AP. Considering the fact that the entropy differences are likely to be small in AP and Me-AP, as in the case of MeO-AP, the present results are consistent with previous studies, which predict that the energy levels of the lowest np* and pp* states of AP and Me-AP lie closer than those of MeO-AP.[7, 9]
Time-resolved mid-infrared spectra at 0–1 ms of photo-excited acetophenone derivatives in carbon tetrachloride are shown for the 900–1900 cm1 region in Figure 5. The
time-resolved near-infrared spectra of acetophenone derivatives in carbon tetrachloride (not shown) are very similar to those in benzene, thus showing that the T1and T2energy gaps in the
two solvents do not differ significantly. We used carbon tetrachloride for the mid-infrared measurements because of much wider spectral windows. In these spectra, the negative peaks represent the depletion of the ground state and the positive peaks the generation of the excited triplet state. The spectra of CF3-AP and CH3O-AP in the T1states have already
been reported by Srivastava et al., and the bands at 1326 and 1462 cm1have been assigned to the CO stretching modes.[16]
In contrast to the prominent vibrational bands observed for CF3-AP and MeO-AP, no apparent positive peaks were
observed for both AP and Me-AP. Such an absence of sharp vibrational bands in the excited triplet states has previously been reported for benzophenone.[14]Since the CO stretching
bands of CF3-AP and MeO-AP have appreciable intensities,
this absence of sharp CO stretching bands in AP, Me-AP, and benzophenone can not be ascribed to their small absorption cross-sections. It is more likely that the vibrational bands of those triplet states are broadened so much for these three molecules that the prominent peaks can not observed. Such broadenings can happen when the exchange between the two triplet states takes place much faster than vibrational
dephasing. The vibrational dephasing time is usually on the order of one picosecond or longer, which corresponds to the vibrational band widths of several wavenumbers. If the exchange between the two triplet states occurs faster than one picosecond, then the effective vibrational dephasing time becomes much shorter than the intrinsic vibrational dephas-ing time and the resultant vibrational band width is much larger than usual. Although the mechanism of this fast exchange is yet to be elucidated, it is highly likely that the perturbations caused by the surrounding solvent molecules play an essential role and that the rate of exchange becomes larger as the two triplet states come closer together. The fast exchange may well be effective for AP, Me-AP, and benzo-phenone but not for CF3-AP and MeO-AP. This supposition
of fast exchange is consistent with the idea that the thermal equilibrium is reached between the T1and T2states of AP,
Me-AP, and MeO-AP.
In conclusion, the thermal equilibrium with fast exchange between the T1and T2states elucidated in the present study
very well accounts for the substituent dependence of photo-reduction activity of acetophenone derivatives as summarized in the following three points: 1) The population distribution between the T1and T2states is reflected in the intensity ratio
of the lower- and higher-wavenumber components in the triplet–triplet absorption spectra. The population is deter-mined by the free energy difference that is almost equal to the determined enthalpy difference because of very small entropy difference. 2) Only the np* state has intrinsic photoreduction activity. This conclusion is consistent with the fact that the maximum degree of mixing (b2) of the np* state into the
lowest pp* states is only approximately 5 %.[11] 3) The
apparent photoreduction activity is expressed by the product of the intrinsic reaction rate of the np* states and its population. There is no need to consider the intrinsic reactivity of the pp* state. Thermal equilibrium between the np* and pp* states plays an important role for acetophe-none, 4’-methylacetophenone and 4’-methoxyacetopheacetophe-none, which all show slightly different photoreduction character-istics.
Received: July 26, 2010
Published online: October 15, 2010
.
Keywords: IR spectroscopy · ketones · photochemistry · time-resolved spectroscopy · triplet excited states[1] N. J. Turro, Modern Molecular Photochemistry, University Science Books, California, 1978.
[2] M. A. El-Sayed, R. J. Leyerle, Chem. Phys. 1975, 62, 1579. [3] S. K. Lower, M. A. El-Sayed, Chem. Rev. 1966, 66, 199. [4] J. C. Scaiano, J. Photochem. 1973/1974, 2, 81.
[5] G. S. Hammond, W. M. Moore, J. Am. Chem. Soc. 1959, 81, 6334; W. M. Moore, G. S. Hammond, R. P. Foss, J. Am. Chem. Soc. 1961, 83, 2789; G. S. Hammond, W. P. Baker, W. M. Moore, J. Am. Chem. Soc. 1961, 83, 2795.
[6] G. Porter, P. Suppan, Trans. Faradey Soc. 1965, 61, 1664. [7] H. Lutz, M. C. Duval, E. Breheret, L. Lindqvist, J. Phys. Chem.
1972, 76, 821; H. Lutz, E. Breheret, L. Lindqvist, J. Phys. Chem. 1973, 77, 1758.
Figure 5. Time-resolved infrared vibrational spectra at 0–1 ms of photo-excited a) CF3-AP, b) AP, c) Me-AP, and d) MeO-AP.
[8] N. C. Yang, D. S. McClure, S. Murov, J. J. Houser, R. Dusenbery, J. Am. Chem. Soc. 1967, 89, 5466; N. C. Yang, R. L. Dusenbery, J. Am. Chem. Soc. 1968, 90, 5899.
[9] D. R. Kearns, W. A. Case, J. Am. Chem. Soc. 1966, 88, 5087. [10] P. J. Wagner, M. J. May, A. Haug, D. R. Graber, J. Am. Chem.
Soc. 1970, 92, 5269; P. J. Wagner, A. E. Kemppainen, H. N. Schott, J. Am. Chem. Soc. 1973, 95, 5604.
[11] S. W. Mao, N. Hirota, Mol. Phys. 1974, 27, 327; E. T. Harrigan, N. Hirota, Mol. Phys. 1976, 31, 663.
[12] M. Berger, E. McAlpine, C. Steel, J. Am. Chem. Soc. 1978, 100, 5147.
[13] See the Supporting Information.
[14] S. Yabumoto, S. Sato, H. Hamaguchi, Chem. Phys. Lett. 2005, 416, 100.
[15] F. Bergmann, Y. Hirshberg, J. Am. Chem. Soc. 1943, 65, 1429; G. S. Hammond, P. A. Leermakers, J. Am. Chem. Soc. 1962, 84, 207.
[16] S. Srivastava, E. Yourd, J. P. Toscano, J. Am. Chem. Soc. 1998, 120, 6173.
