Photochemistry and photodissociation of benzosultine and
naphthosultine: electronic relaxation of sultines and kinetics and
theoretical studies of fragment o-quinodimethanes
Kuo-Chun Tang
a, Sheng-Jui Lee
a, San-hui Chi
a, Kuen-Ling Lu
a, Wei-Chen Chen
a,
Chin-hui Yu
a, I-Chia Chen
a,∗, Shu-Li Wu
b, Chun-Cing Chen
b, Wen-Dar Liu
b,
Liang-Jyi Chen
b, Niann S. Wang
b,∗∗, Wen-Sheng Chung
b,∗∗aDepartment of Chemistry, National Tsing Hua University, 101 Kuang Fu Road Sec. 2, Hsinchu, Taiwan 30013, ROC bDepartment of Applied Chemistry, National Chiao Tung University, Hsinchu, Taiwan 30050, ROC
Received 26 March 2004; received in revised form 3 August 2004; accepted 4 August 2004 Available online 11 September 2004
Abstract
Fluorescence decays of benzosultine and naphthosultine excited at 263 nm are detected with time-correlated single-photon counting (TC-SPC). Both molecules display biexponential decay with a rapid component (time constants 90 and 350 ps for benzosultine and naphthosultine, respectively) assigned to an electronically excited state Snat high energy and the slow one (constants 7.5 and 9.0 ns for benzosultine and naphthosultine, respectively) to the S1state. Dissociation of both molecules was investigated with nanosecond laser flash photolysis combined with transient absorption to detect the intermediates and products. With excitation at 266 nm, benzosultine yields a transient with absorp-tion maximum atλmax = 370 nm; this transient has a short-lived component with lifetime around 1s and a long-lived component. Both components are insensitive to molecular oxygen. The short-lived component is tentatively assigned to the dissociation intermediate and the long-lived to be singlet o-benzoquinodimethane (o-BQDM). Photolysis of naphthosultine yields two transient species with absorption atλmax = 420 and 520 nm; we assign the former band to triplet–triplet absorption of naphthosultine and the latter to absorption by the product singlet
o-naphthoquinodimethane. Optimal geometries, energetics, and vertical transitions of benzosultine, naphthosultine, o-benzoquinodimethane,
and o-naphthoquinodimethane are calculated using methods based on density-functional theory (B3LYP/6-311++G**) and time-dependent density-functional theory (TD-DFT). The results of these calculations imply that the ground electronic state of these two o-quinodimethanes (o-QDM) is singlet with a structure of diene form. Their triplet states display a biradical structure. The energy separation between singlet and triplet states of o-benzoquinodimethane is calculated to be ca. 93.6 kJ/mol, but for o-naphthoquinodimethane, it is only ca. 27.8 kJ/mol. © 2004 Elsevier B.V. All rights reserved.
Keywords: Benzosultine; Naphthosultine; o-Quinodimethane; Photochemistry
1. Introduction
o-Quinodimethane (o-QDM, also known as o-xylylene) has attracted much attention of both physical and synthetic chemists and it is known as a versatile building block in or-ganic synthesis [1–7]. The major synthetic application of
∗Corresponding author. Tel.: +886 35715131; fax: +886 35721614. ∗∗Co-corresponding author.
E-mail address: [email protected] (I.-C. Chen).
o-QDM is its cycloaddition reactions with dienophiles to form multicyclic rings. Cava et al. [2–4] were the first to propose a quinonoid bromide (dibromo-o-QDM) as an in-termediate in the reaction of␣,␣,␣,␣-tetrabromo-o-xylene with iodide ion. Segura and Mart´ın reviewed the structures, properties, and generations of o-QDM as reaction inter-mediates and their applications in synthetic, material, and polymer chemistry[1]. In this work, we shall focus on the spectral properties and generation of o-QDMs using flash photolysis.
1010-6030/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jphotochem.2004.08.004
Flynn and Michl[5,6]photolyzed 1,4-dihydrophthalazine in a rigid glass and identified an intermediate, produced after photolysis, with absorption band atλmax= 373 nm to be o-benzoquinodimethane (o-BQDM) 1. Wintgens et al.[7a] pho-tolyzed 1,1,3,3-tetramethyl-2-indanone in benzene to pro-duce tetramethyl-o-BQDM; they assigned an absorption at
λmax≈ 358 nm and a long lifetime 9.1 min to the singlet form,
from its low reactivity with oxygen. A short-lived intermedi-ate absorbing atλ ≈ 320 nm with lifetime 580 ns was assigned to a triplet biradical form of tetramethyl-o-BQDM. Product tetramethyl-o-BQDM underwent slow antarafacial 1,5-hydrogen shift to form 1-isopropenyl-2-isopropylbenzene, hence might be observable by flash photolysis[7b].
The results of Flynn and Michl [5,6] and of Wintgens et al. [7a] show the generation of o-BQDM from 1,4-dihydrophthalazine or indanone that undergoes either a retro-Diels-Alder or a cheletropic reaction to lose nitrogen or carbon monoxide. Generation of o-BQDM from sultines and sulfolenes has been reported[8–11]; however, we prefer the former precursors because of their thermal stability, mild reaction temperatures, and photochemially active as opposed to the sulfolenes. For example, o-BQDM can be obtained from sultines at temperatures as low as 80◦C whereas for sulfolenes, temperature greater than 280◦C is required. Durst and co-workers [8–11]proposed that benzosultine 2 dissociates to yield either o-BQDM 1 or biradical 3 under either thermolysis conditions (>80◦C) or 254-nm irradiation (see Scheme 1). Although their work has been performed on dissociation of sultines, the dissociation mechanism for forming o-BQDM 1 or biradical intermediate 3 is not yet understood. Moreover, identification of the existence of biradical 3 after photodissociation has not been reported.
We have been interested in the studies of heterocyclic fused sultines and their application in Diels-Alder reactions with electron-poor olefins and [60]fullerenes[12]. Fullerenes include many other homologues such as C70 and C84 etc. Since our trapping study involves only [60]fullerene, there-fore, the [60] should be kept to clarify our work. In the present work, we choose the two simplest sultines—benzosultine 2 and naphthosultine 5—and employ methods of laser flash photolysis and time-correlated single-photon counting (TC-SPC) to investigate their reactions and the fluorescence decay
Scheme 1.
Scheme 2.
of excited states before dissociation. A possible reaction of naphthosultine[12d]is shown inScheme 2.
Transient absorption spectra of these two compounds af-ter photolysis and the quenching reactions of the o-BQDM 1 with alkenes are reported. From kinetic studies of the transient species with singlet and triplet quenchers, one can elucidate the electronic structure and spin states of these species. Fur-thermore, we employ density functional methods[13–15]to calculate the structures and nature of electronically excited states of sultines and o-QDMs. The results of these calcula-tions yield estimates of energetics among these compounds and assist in identification of intermediates. Time-dependent density-functional theory (TD-DFT) [16–19] is applied to compute transition energetics and oscillator strengths. The calculated vertical excitation energies and geometries are es-sential in explaining the mechanism of dissociation.
2. Experimental
2.1. Materials
Benzosultine[9]and naphthosultine[12d]were prepared according to the method described by Hoey and Dittmer[20].
2.2. Measurements
Nanosecond pump-probe transient-absorption experi-ments were performed using a pulsed xenon flash lamp (75 W) as the probe pulse. The fourth harmonic of output from a Nd:YAG laser (Continuum NY60B) at 266 nm with pulse width 10 ns served for photolysis; this beam was gently focused with a cylindrical lens to a rectangular size 10 mm× 2 mm on a quartz sample cell (length 1 cm). For some early tests, the beam from a KrF excimer laser (Lambda Physik, Lextra 50, 248 nm) was used. The experimental data obtained from both pump lasers are similar except that decay traces ob-tained with the YAG laser are superior in the short-time region because of less RF interference. The white-light probe beam passing through the sample cell was focused with two lenses onto the entrance slit of a monochromator (Acton SP-300I). Both pump (10–20 mJ/pulse) and probe pulses were crossed perpendicularly; the overlap distance was 1 cm. An inten-sified charge-coupled device (ICCD, Princeton Instrument, Model 576G/RB) served to detect the transient absorption spectra. The background was subtracted from signals col-lected from 20 laser shots; the remaining signals were then averaged to yield the spectra.
Decays of transient species were detected at a fixed wave-length by a photomultiplier tube (EMI 9829QB, response time 5 ns) mounted at another exit of the monochromator. The signal was averaged with a digital oscilloscope (LeCroy 9450) for about 20 laser shots. Each sample in acetonitrile or hexane with absorbance 0.3–0.5 at excitation wavelength 266 or 248 nm was degassed to remove oxygen before use. After 20 laser shots, each solution was discarded except for those retained for study of end products. In each case, the decay kinetics, fitted to a first-order equation, yield the experimen-tal rate constants. Measurements of reactivity of transients with molecular oxygen and alkenes (including fumaronitrile (FN) and N-phenylmaleimide (NPM)) were performed. All measurements were performed at 298± 2 K.
Picosecond lifetimes were measured with time-correlated single-photon counting. The third harmonic of output from a femtosecond Ti:sapphire amplifier (1 kHz, Spectra Physics, Hurricane) at 263 nm served as the pump beam (10J/pulse). The temporal resolution (≈120 ps) is limited by combination of the response of MCP-PMT (Hamamatsu R3809U-50) and the time-to-amplitude converter (EG&G ORTEC). We used benzosultine in acetonitrile at concentration≤5 × 10−5M to prevent self-quenching. Fluorescence decays were fitted to exponential functions with convolution over the response function of the detection system.
Fluorescence quantum yields ΦF of benzosultine and naphthosultine were measured using benzene in cyclohex-ane (ΦF= 0.07) as the standard; they areΦF= 0.09 and 9× 10−4for benzosultine and naphthosultine, respectively. 2.3. Computational methods
Hybrid three-parameter Becke exchange functionals and Lee–Yang–Parr correlation functionals (B3LYP)[21–26], as well as B3P86[27], and B3PW91[28–31]were used to obtain optimized geometries of o-BQDM. Harmonic vibrational fre-quencies are calculated at an optimized geometry and applied to obtain the zero-point energy. The geometries obtained with these three methods vary within 3%; hence, for calculations on other molecules, only method B3LYP with basis set up to 6-311++G** was sufficient to achieve reliable results. Ver-tical transitions to electronically excited singlet and triplet states of o-QDM and sultines were calculated for these opti-mal geometries with TD-DFT. Both DFT and TD-DFT cal-culations were performed with GAUSSIAN 98 program[32]
(on a SGI Origin workstation and personal computers). Ana-lytic first derivatives were utilized in geometry optimization, and vibrational frequencies were calculated analytically at each stationary point.
3. Results
3.1. Benzosultine 2
The stationary UV spectrum of benzosultine in acetonitrile displays a weak absorption with maximum at 260 nm and a
Fig. 1. Steady-state UV–vis (solid line) and dispersed fluorescence (dashed line) spectra of (a) benzosultine in acetonitrile concentration 5.8× 10−4M and (b) naphthosultine, 3.3× 10−5M. Dispersed fluorescence spectra were recorded at excitation wavelength 266 nm.
strong absorption at 235 nm. For 266-nm excitation, the flu-orescence spectrum consists of a band at λmax ≈ 280 nm. These absorption and fluorescence emission spectra are dis-played inFig. 1a. InFig. 2a, the fluorescence decay curve measured with TCSPC at 263-nm excitation is shown; this curve displays biexponential decay. We fitted this curve to a biexponential function and the best-fit time constants are 90
± 10 ps and 7.5 ± 0.2 ns for rapid and slow components,
re-spectively. The relative amplitudes of these two components vary with detection wavelength, with an increased slow com-ponent when the long-wavelength region is detected.
Upon 266-nm photolysis, transient absorption spectra of benzosultine in solution were measured; they display a broad absorption atλmax= 370 nm, as shown inFig. 3; this transient decays with time constant 1.04± 0.05 s in acetonitrile so-lution and 1.15± 0.05 s in hexane to reach a plateau lasting for 1 ms (limited by the pulse duration of the probe beam). These decays show low reactivity to molecular oxygen and independent of concentration (1–5× 10−3M) of benzosul-tine. The short component is quenched by NPM at rate con-stant (2.1 ± 1.0) × 108M−1s−1and FN at (8.5± 4.2) ×
Fig. 2. Fluorescence decay of (a) benzosultine (2.3× 10−5M) and (b) naph-thosultine (5.3× 10−5M) monitored at 320 nm in acetonitrile at 263-nm ex-citation with fitted time constantsτfandτsas indicated for the fast and slow
components, respectively. Deconvolution from the system response function for an assumed Gaussian function with FWHM = 120 ps is performed for all curves.
107M−1s−1. This transient species appeared promptly after photolysis, more quickly than instrumental resolution (10 ns), and then underwent some decay process.
When benzosultine 2 in acetonitrile was irradiated with 254 nm light from a Hg lamp, SO2was extruded and
moder-Fig. 3. Transient absorption spectrum obtained for benzosultine in acetoni-trile at 266-nm excitation. Inset: absorption curve measured at 370 nm decays at time constantτ = 1.04 ± 0.05 s to a plateau region.
Table 1
Results of benzosultine 2 and benzosulfolene 4 with olefins in acetonitrile solutions after 254-nm photolysisa
Entry Diene precursor
Olefin Product (%) Yield (%)b
4 Diels-Alder adducts 1 2 None 52 0 52 2 2 NPM 7 89 96 3 2 FN 27 17 44 4 2 DMF 38 18 56 5 2 DMAD 16 49 65 6 2 DPA No reaction 0 7 4 None No reaction 0 8 4 FN No reaction 0
a DMAD is dimethylacetylenedicarboxylate; DMF is dimethylfumarate;
NPM is N-phenylmaleimide; FN is fumaronitrile; DPA is diphenylacetylene. All the olefins used in the reaction were 1.5 equivalent vs. sultine or sulfolene.
b Yields determined by1H NMR and corrected for recovered starting
ma-terials.
ate to excellent yields of Diels-Alder adducts were obtained in the presence of 1.5 equivalent of electron-poor olefins, such as N-phenylmaleimide, dimethylfumarate (DMF), dimethy-lacetylenedicarboxylate (DMAD), and fumaronitrile. The yields of production are in the range of 44–96% and the re-sults are summarized inTable 1. Furthermore, when the above solutions of 2 with olefins and triplet sensitizers[33a](such as, acetone (ET=∼78 kcal/mol), benzil (ET=∼54 kcal/mol),
or benzophenone (ET=∼69 kcal/mol)) were irradiated with
300 nm light, only benzosultine 2 was recovered and no Diels-Alder adducts between o-BQDM and olefins were found; however, the expected oxetane products between acetone and fumaronitrile were found[33b]. The results imply that the triplet energy of these sensitizers is not sufficient to transfer to the triplet of benzosultine.
3.2. Naphthosultine 5
The synthesis and sealed tube reaction of naphthosultine
5 in the presence of various dienophiles have been reported
elsewhere[33c]. Similar to benzosultine, the stationary ab-sorption spectrum displays abab-sorption with maxima at 280 and 220 nm, as shown inFig. 1b. At 266-nm excitation, naph-thosultine displays much less fluorescent intensity than ben-zosultine. The fluorescence emission spectrum from a recrys-tallized sample of 5 in acetonitrile displays two features at 340 and 400 nm. From measurement of fluorescence excita-tion spectra at these two bands, the band at 400 nm has an excitation spectrum different from absorption of naphthosul-tine, implying that this band results from emission of an im-purity. Because the fluorescence efficiency of naphthosultine
5 is low, an impurity in only trace proportion but with
mod-erately efficient emission would appear in the fluorescence spectra. The excitation spectrum obtained on monitoring the band at 340 nm agrees with the absorption of naphthosultine; therefore, this band is attributed to emission of naphthosul-tine.
Fig. 4. Transient absorption spectrum obtained for naphthosultine in ace-tonitrile at 266-nm excitation. Inset: decay curves of absorption measured at 420 nm (dashed line) and at 520 nm (solid line) at time constantsτ = 1.6
± 0.1 s and 1.8 ± 0.1 s, respectively.
The emission curves monitored at 320 nm show biexpo-nential decay with fitted time constants 330 ps and 8.5 ns, but the impurity displays time constants 700 ps and 3.0 ns when monitored at 410 nm. The identity for this impurity is not known. The TCSPC curve for naphthosultine with the best exponential fit is displayed inFig. 2b.
At 266-nm photolysis, the transient absorption spectrum for naphthosultine in acetonitrile is displayed inFig. 4; this spectrum shows three absorption features centered at 310, 420, and 520 nm, decaying with time constants 1.3, 1.8, and 1.6s, respectively. About 300 nm near the edge of emission from a Xe lamp, the uncertainty in absorbance is relatively large; theλmaxposition is accordingly limited by the white-light source. When the naphthosultine solution was irradiated over 20 shots, transient absorption at 420 nm increased and the feature at 520 nm decreased. The absorption at 420 nm appeared at long radiation even without oxygen; this product might be naphthocyclobutene, which was produced from 6. For fresh solutions, the absorption decay of transient species at 420 and 520 nm is shown as an inset inFig. 4.
The species at 420 nm is quenched by molecular oxy-gen (rate constant (8.9± 0.7) × 108M−1s−1, as shown in
Fig. 5), and is quenched by FN (rate constant (3.9± 0.7)
× 108M−1s−1). The species absorbing at 520 nm is
insensi-tive to quenching with oxygen but is quenched by FN at rate constant (3.5± 0.8) × 108M−1s−1.
3.3. Geometries and energetics
Two geometric isomers 2a and 2b separated by 10.3 kJ/mol are obtained for benzosultine using B3LYP/6-311++G**, one corresponding to an axial SO bond and the other an equatorial one in the sulfinic ester ring. These opti-mal geometries are displayed inFig. 6with bond distances. For such small energy difference of these two isomers and no experimental evidence, we assume that both structures are
Fig. 5. Kinetic measurement of short-lived transient at 420 nm produced from naphthosultine after 266-nm excitation. The quencher is O2; kII= (8.9 ± 0.7) × 108M−1s−1.
equally probable. Vertical transitions to singlet and triplet excited states at wavelengths greater than 200 nm using TD-DFT based on both optimal geometries for their ground elec-tronic states are listed in Table 2. Accordingly, bands ob-served at 260 and 235 nm are assigned to be S1−S0 and S4−S0/S5−S0transitions, respectively, with large oscillator strengths from the results of calculations. From inspection of molecular orbitals on both structures, the lowest transition has dominant–* character. Five triplet states are present below the position of S1−S0, of which some might be re-sponsible for weak absorption extending to 320 nm. The op-timal geometry of the lowest triplet state of benzosultine is calculated, with its vertical transitions to other triplet states. According to these calculations, strong triplet–triplet absorp-tions occur at 286 nm for structure 2a and 308 and 369 nm for structure 2b. Because the observed transient about 370 nm is insensitive to molecular oxygen, it is not expected to arise from triplet–triplet absorption of the precursor.
Isomerization from benzosultine to benzosulfolene ex-hibits an energy barrier around 254 kJ/mol (with basis set 3-21G*); thermal conversion between these two molecules is hence unlikely. In calculating the dissociation pathways, we assumed first that only one chemical bond is cleaved. One converged structure in singlet and the other in triplet are attained and their optimal structures are displayed in
Fig. 6. The triplet intermediate lies ∼170 kJ/mol above the zero point of singlet benzosultine 2a, and has vertical T–T transitions at 560, 340, and 267 nm with oscillator strengths about 0.0311, 0.0588, and 0.1212, respectively. The singlet intermediate lies about 192 kJ/mol with vertical transitions at 406 and 386 nm and oscillator strengths about 0.0128 and 0.0266 (at level B3LYP/6-311++G**), respectively. Because we observed no absorption in the 560-nm region, this is un-likely to correspond to transient absorption observed here at 370 nm. Besides, a triplet intermediate is expected to react ef-ficiently with oxygen. Reaction for the singlet intermediate to o-BQDM + SO2has a small barrier around 4 kJ/mol. Hence,
under room temperature, this intermediate should proceed immediately to products after it is formed. The lowest triplet
Fig. 6. Optimized geometries (in ˚A) for two equatorial and axial geometric isomers of benzosultine, triplet benzosultine, singlet and triplet intermediate, and benzosulfolene at B3LYP/6-311++G**.
Table 2
Results of TD-DFT on vertical electronic transitions of benzosultine using method B3LYP/6-311++G**, with calculated oscillator strengths listed in parentheses
B3LYP/6-311++G** Experiment Benzosultine 2a; Sn−S0;λ (nm) Benzosultine 2b; Sn−S0;λ (nm) Benzosultine 2a; Tn−T1;λ (nm) Benzosultine 2b; Tn−T1;λ (nm) Benzosulfolene 4; Sn−S0;λ (nm) Benzosultine; λ (nm) 244 (0.0110) 251 (0.0149) 779 (0.0006) 1042 (0.0002) 239 (0.0041) 260 241 (0.0033) 246 (0.0008) 531 (0.0060) 789 (0.0023) 215 (0.0003) 228 (0.0024) 237 (0.0002) 510 (0.0114) 694 (0.0054) 205 (0.0084) 223 (0.0129) 232 (0.0359) 397 (0.0009) 535 (0.0018) 201 (0.1070) 235 219 (0.0173) 217 (0.0131) 382 (0.0040) 491 (0.0031) 210 (0.0049) 210 (0.0144) 370 (0.0027) 430 (0.0020) 207 (0.0264) 209 (0.0096) 342 (0.0000) 428 (0.0004) 206 (0.0418) 203 (0.0762) 339 (0.0004) 418 (0.0104) 203 (0.0030) 203 (0.0090) 330 (0.0025) 400 (0.0037) 198 (0.0146) 199 (0.0085) 324 (0.0039) 369 (0.0203) 307 (0.0098) 362 (0.0069) 293 (0.0144) 347 (0.0062) 286 (0.0349) 343 (0.0027) 281 (0.0090) 340 (0.0001) 278 (0.0062) 334 (0.0011) 277 (0.0007) 308 (0.0673) 272 (0.0125) 297 (0.0040) 269 (0.0142) 296 (0.0034) 264 (0.0015) 293 (0.0005) 256 (0.0089) 274 (0.0023) Tn−S0 Tn−S0 Tn−S0 337 338 335 279 294 275 275 277 272 274 276 250 253 256 208 229 245 205 229 236 200 227 228 220 222 210 210 207 206 207 204 204 204 202 202
state of benzosultine lies about∼339 kJ/mol above the zero point of 2a; its optimized structure appears also inFig. 6.
The optimized structures of singlet and triplet naphthosul-tine with energy difference∼251 kJ/mol are shown inFig. 7. Similar to that for benzosultine, two geometric isomers 5a and
5b for naphthosultine separated by 10.4 kJ/mol are obtained.
Isomerization to naphthosulfolene on the singlet electronic ground surface has a calculated energy barrier 257 kJ/mol, also similar to that obtained for benzosultine at the same level of calculation. Results for vertical transition energies calcu-lated for triplet states of naphthosultine are summarized in
Table 3.
The optimized structures of o-QDMs for both the lowest singlet and triplet states are calculated and are displayed in
Fig. 8. Point groups for optimal geometries are C2and C2v
for the singlet and triplet ground states, respectively. Simi-lar structures were obtained with all three density-functional methods and bond angles and distances calculated lie within 1–2%. With the symmetry restriction relaxed to C2, all DFT
methods yield twisted geometries for the singlet state o-BQDM with a dihedral angle 15.7–18.5◦, but the energies are nearly the same when the structures are restricted to be planar (C2v). In calculations on1o-BQDM (left superscript denotes
a singlet state), Sakai[34]reported an optimized planar ge-ometry with CASSCF/6-31G**, but a twisted gege-ometry with DFT, which resemble our results. Based on the low frequency for the twisting vibration, 1o-BQDM is expected to have a nearly planar Kekule geometry with a floppy conformation along the twisting motion (the frequency for the twisting vi-bration is ∼81 cm−1). The optimized structure (B3LYP/6-311++G**) for the lowest triplet state of o-BQDM is planar with carbon–carbon bonds of length 1.38–1.41 ˚A except one bond 1.453 ˚A, implying delocalization of electrons from the C C bond connecting to the benzene ring to around the benzene ring. This result indicates that the triplet exhibits a biradical character. This triplet state lies≈93.6 kJ/mol above its singlet. From results of theoretical calculations, we con-clude that o-BQDM has a singlet ground state.
Fig. 7. Optimized geometries (in ˚A) for equatorial and axial geometric isomers of singlet and triplet naphthosultine at B3LYP/6-311++G**.
Table 3
Results of TD-DFT on vertical electronic transitions of naphthosultine, with calculated oscillator strengths listed in parentheses
B3LYP/6-311++G** Experiment
Naphthosultine 5a Naphthosultine 5b Naphthosultine 5a Naphthosultine 5b Naphthosultine
Sn−S0;λ (nm) Tn−S0;λ (nm) Sn−S0;λ (nm) Tn−S0;λ (nm) Tn−T1;λ (nm) Tn−T1;λ (nm) Sn−S0;λ (nm) Tn−T1;λ (nm) 294 (0.0428) 463 297 (0.0396) 466 856 (0.0001) 862 (0.0000) 260 289 (0.0017) 326 290 (0.0020) 327 774 (0.0004) 774 (0.0000) 260 (0.0212) 303 271 (0.0126) 304 658 (0.0025) 657 (0.0004) 246 (0.0642) 300 251 (0.0431) 302 630 (0.0143) 584 (0.0046) 240 (0.0141) 290 238 (0.0036) 295 470 (0.0040) 483 (0.0006) 232 (0.0063) 280 233 (0.2686) 291 441 (0.0065) 449 (0.0078) 235 230 (0.0002) 260 232 (0.0215) 272 431 (0.1734) 431 (0.1575) 420 224 (0.3696) 244 230 (0.0010) 242 426 (0.0077) 426 (0.0192) 223 (0.0058) 239 221 (0.8361) 240 405 (0.0003) 415 (0.0001) 221 (0.4294) 233 221 (0.0672) 236 391 (0.0004) 398 (0.0014) 220 (0.2074) 230 218 (0.1092) 232 376 (0.0000) 359 (0.0003) 217 (0.2737) 229 216 (0.0476) 231 349 (0.0076) 356 (0.0050) 210 (0.0511) 227 210 (0.0660) 229 341 (0.0004) 354 (0.0007) 208 (0.0831) 224 209 (0.0200) 227 335 (0.0012) 342 (0.0023) 207 (0.0025) 220 209 (0.0125) 222 328 (0.0021) 333 (0.0020) 206 (0.0086) 219 207 (0.0206) 218 320 (0.0109) 318 (0.0327) 204 (0.0501) 216 206 (0.1170) 216 311 (0.0286) 309 (0.0033) 204 (0.0325) 212 204 (0.0128) 212 308 (0.0065) 308 (0.0090) 203 (0.0058) 210 203 (0.0124) 211 300 (0.0154) 302 (0.0075) 201 (0.0141) 208 201 (0.0183) 209 295 (0.0014) 294 (0.0009) 292 (0.0001) 288 (0.0011) 285 (0.0003) 285 (0.0041) 285 (0.0012) 284 (0.0019) 281 (0.0137) 279 (0.0134) 276 (0.0206) 278 (0.0012)
Fig. 8. Calculated geometries (B3LYP/6-311++G**) of the ground state of (a)1o-BQDM, (b)3o-BQDM, (c) benzocyclobutene, (d)1o-NQDM, and (e)
3o-NQDM. Bond lengths are indicated in ˚A.
Based on optimized geometries, the vertical transitions of o-BQDM were calculated with TD-DFT; the first 10 transi-tions for both singlet and triplet states are listed inTable 4. For
1o-BQDM (C
2symmetry), the transition with greatest
oscil-lator strength lies at 402 nm, and the next high energy lies at 240 nm with a modest oscillator strength.3o-BQDM has a strong absorption at 345 nm with oscillator strength 0.1173 and a high-energy band at 295 nm with oscillator strength 0.1706. With the position shift from methylation taken into account, these results of theoretical calculations agree with assignments that Wintgens et al.[7a]made on triplet tetram-ethyl o-BQDM.
o-BQDM is known to react further to form spiro dimers even at low temperatures[35]. Dimers with four geometric isomers are calculated: chair, boat, twisted boat, and spiro with the former three forms lying at comparable and lower
energies; their optimal structures are shown inFig. 9. The first 10 singlet vertical transitions of these dimers calculated with TD-DFT are also listed with oscillator strengths inTable 4. The spiro dimer shows a strong absorption at 321 nm and the other dimers have strong absorption near 220 nm. According to the results of experiments, the rise of the transient species is rapid and independent of sample concentration and laser power; hence, no dimer is suitable for attribution of the tran-sient at 370 nm.
In calculating geometry and energy of o-NQDM, we used method B3LYP with basis set up to 6-311++G**. Similar to the results obtained for o-BQDM, the singlet ground state of o-NQDM has an optimized non-planar geometry (point group C2) with a dihedral angle 23.9◦but its energy
is within 1.0 kJ/mol of a planar structure (C2v). Hence,
Table 4
Results of TD-DFT using method B3LYP on o-BQDM, singlet benzocyclobutene, and o-BQDM dimers for their first 10 singlet vertical electronic transitions, with calculated oscillator strength listed in parentheses
B3LYP/6-311++G** 1o-BQDM (λ/nm) 3o-BQDM (λ/nm) Benzocyclobutene (λ/nm) Dimer (chair) (λ/nm) Dimer (boat) (λ/nm) Dimer (twist) (λ/nm) Dimer (spiro) (λ/nm) 402 (0.1177) 486 (0.0002) 231 (0.0232) 240.9 (0.0033) 242.1 (0.0069) 239.3 (0.0015) 321.2 (0.1704) 268 (0.0026) 427 (0.0025) 201 (0.0005) 239.8 (0.0000) 240.7 (0.0039) 239.3 (0.0000) 278.2 (0.0086) 265 (0.0016) 345 (0.1173) 177 (0.4134) 230.4 (0.0000) 223.5 (0.0029) 223.8 (0.0024) 268.0 (0.0011) 250 (0.0044) 324 (0.0001) 175 (0.7589) 221.6 (0.0000) 223.1 (0.0039) 222.7 (0.0019) 264.8 (0.0022) 248 (0.0002) 317 (0.0107) 174 (0.0042) 219.6 (0.0094) 221.2 (0.0154) 218.7 (0.0046) 259.8 (0.0037) 247 (0.0017) 296 (0.0035) 172 (0.0012) 218.6 (0.0438) 220.3 (0.0054) 216.9 (0.0000) 253.2 (0.0023) 240 (0.0190) 295 (0.1706) 171 (0.0116) 214.5 (0.0000) 219.8 (0.0233) 216.8 (0.0098) 244.5 (0.0173) 219 (0.0215) 292 (0.0000) 167 (0.0000) 214.5 (0.0024) 219.7 (0.0098) 216.2 (0.0137) 241.1 (0.0104) 218 (0.0025) 277 (0.0268) 167 (0.0000) 212.8 (0.0440) 218.3 (0.0145) 213.6 (0.0043) 239.6 (0.0099) 212 (0.0016) 260 (0.0021) 165 (0.0000) 210.5 (0.0000) 217.7 (0.0109) 213.3 (0.0000) 238.3 (0.0005)
corresponding to twisting motion is expected to have only a shallow minimum. The lowest triplet state of o-NQDM is only 27.8 kJ/mol above its singlet state, much less than the separation calculated for o-BQDM; for this reason, mixing between the singlet and triplet o-NQDM is expected. 3 o-NQDM has a planar geometry and biradicaloid nature with delocalization of electrons around the naphthalene ring.
The TD-DFT results on o-NQDM are listed in Table 5. According to the results of calculations, a strong absorption band would appear at 608 nm (C2symmetry) for the singlet
state with oscillator strength 0.07; vertical T–T transitions
Fig. 9. Calculated geometries (B3LYP/6-311++G**) of the ground states of o-BQDM dimers in its (a) chair, (b) boat, (c) twisted boat, and (d) spiro forms.
for the triplet state occur at 388 and 313 nm with oscillator strengths 0.2209 and 0.0832, respectively.
4. Discussion
4.1. Benzosultine
With excitation at 263 nm from femtosecond laser pulses, possibly both S1and Sn(n = 2–5) are accessed. From
biexponen-Table 5
Calculated vertical excitation energies and oscillator strength of o-NQDM using TD/B3LYP/6-311++G**
B3LYP/6-311++G** This work
(λ (nm)) Gisin and Wirza(λ (nm)) 1o-NQDM 3o-NQDM λ (nm) Oscillator strength λ (nm) Oscillator strength 607.6 0.0692 565.9 0.0007 520 540 351.7 0.0014 456.4 0.0008 311.6 0.0138 393.0 0.0014 305.0 0.0059 388.4 0.2209 294.1 0.0022 351.3 0.0006 280.3 0.0002 339.1 0.0117 279.0 0.0024 317.0 0.0366 278.3 0.0029 316.4 0.0004 250.2 0.0044 313.2 0.0036 249.3 0.0014 312.9 0.0832 247.5 0.0057 299.3 0.0073 238.3 1.2960 298.4 0.0000 231.4 0.0015 296.6 0.0097 227.3 0.0218 266.9 0.0043 223.5 0.0091 264.8 0.0000 221.9 0.0965 264.6 0.0001 215.9 0.0238 260.2 0.0000 213.4 0.0157 257.1 0.0001 212.2 0.1351 255.4 1.0050 206.1 0.0326 252.7 0.0000
a Gisin and Wirz, reference[38].
tial fluorescence decay to state Sn and the slow component to S1. The high-energy Snstate can convert internally to any
low-lying electronic states efficiently to display a short life-time, 90 ps. Combining the measured fluorescence quantum yield and the assigned lifetimeτ = 7.5 ns for S1, we estimated the radiative rate constant (=ΦF/τ) to be 1.2 × 107s−1. From the calculated oscillator strength f = 0.0064, we obtained a ra-diative rate constant 0.95× 107s−1, near the value estimated from experimental data.
The observed rise time constant of o-BQDM is less than 10 ns, which is comparable to the decay of the S1 state.
Benzosultine is known to dissociate at temperatures∼80◦C and then undergoes various reactions to form other com-pounds. According to quenching experiments, no triplet tran-sient species is involved within the experimental time scale. For these reasons, we expect that the triplet state is unimpor-tant in this dissociation. Combined the experimental evidence and the results of theoretical calculation, possibly upon pho-toexcitation, sultine molecules undergo rapid conversion to form the singlet intermediate, and then dissociate to prod-ucts. Therefore, we assign the short component in the ob-served transient species absorbing atλmax370 nm to be the dissociating singlet intermediate and the long component is
1o-BQDM. The lifetime of the intermediate is about 1s.
Under pyrolysis, direct dissociation to1o-BQDM + SO2 is
also likely because this pathway lies at lower energy and this product channel is correlated without an exit barrier.
This assignment for1o-BQDM agrees with that made by Wintgens et al.[7a]for product tetramethyl-o-BQDM from
Fig. 10. Schematic diagram for reaction benzosultine→ singlet intermedi-ate→ o-BQDM + SO2. The energy (kJ/mol) shown in here is related to the zero point of benzosultine 2a, calculated with method B3LYP/6-311++G**. Symbol1I and TS denote the singlet intermediate and the transition state,
respectively.
dissociation of 1,1,3,3-tetramethyl-2-indanone; in benzene, this singlet product absorbed at 358 nm. Fujiwara et al.[7c]
also assigned the transient absorption at 370 nm to be sin-glet o-BQDM from 266 nm irradiation of ␣,␣-dichloro or dibromo-o-xylene in cyclohexane solution at room tempera-ture. Compared with the calculated vertical transition 402 nm for 1o-BQDM, the theoretical value yields a deviation of about 8%. However, no triplet o-BQDM, the biradical prod-uct, is observed from benzosultine at 266-nm photolysis.
Flynn and Michl[5,6]found that o-BQDM prepared in a low-temperature matrix underwent rapid dimerization when the temperature of the glass increased. Errede [35] found that o-BQDM underwent ring closure to benzocyclobutene at 300–600◦C, but dimerization to a spiro form at much lower temperatures. According to results of calculations, dimers in chair, boat, and twisted boat forms are energetically more fa-vorable than in spiro form, but the spiro form might have less geometric constraint. Nevertheless, in the present work, the decay of o-BQDM is independent of concentration and too rapid for bimolecular reaction under such low concentration conditions. Cyclization of o-BQDM to benzocyclobutene ex-hibits a high-energy barrier (169 kJ/mol)[36–38]. Ouchi et al.[39–40]used two-color laser photolysis to study the re-actions of o-BQDM. They concluded that benzocyclobutene is a photoproduct of o-QDM and spiro dimer is a thermal product. This explains that under the current experimental condition,1o-BQDM is a relatively long-lived product.
Relaxation of benzosultine after excitation and its corre-lation to products are displayed inFig. 10; the energetics of dimers and benzocyclobutene relative to o-BQDM are shown inFig. 11.
4.2. Naphthosultine
According to the TD-DFT calculations, with excitation at 263 nm, mainly the S8(for 5a) and the S6(for 5b) states are
Fig. 11. Dimerization of o-BQDM and their relative energy calculated with B3LYP/6-311++G**. The listed energy (kJ/mol) is relative to that of o-BQDM.
and corresponds to the onset of absorption to S1. Hence, S1
is the emitting state. We assign the slow component with life-time 9.0 ns in the biexponential fluorescence decay to decay of the S1 state and the rapid component to a high-energy
excited state Sn(n = 2–6 or 8). The measured fluorescence quantum yield is low (ΦF= 0.09%); from the measured life-time for the S1state, we obtained a radiative decay rate
con-stant of 1× 105s−1in contrast to an estimated value of 1.7
× 107s−1from the calculated oscillator strength. Compared
with the fluorescence quantum yields of benzene and naph-thalene in cyclohexane, 0.07 and 0.23, respectively,ΦF of naphthosultine is much smaller. It is possible that the S1state
of naphthosultine is mixed strongly with its triplet state such that the fluorescence intensity is decreased significantly.
Vertical T–T transitions of naphthosultine are calculated to show an absorption band with maximum oscillator strength at 431 nm. We tentatively assign the observed 420-nm band with great absorbance to triplet–triplet absorption of naph-thosultine. This assignment is consistent with the results of TCSPC for singlet–triplet mixing in naphthosultine. How-ever, the calculated vertical transition with dominant oscil-lator strength for triplet biradical product3o-NQDM lies at
388 nm, which deviates about 8% from the observed posi-tion; as this deviation is considered reasonable, one cannot exclude possible formation of 3o-NQDM. Because the en-ergy gap between singlet and triplet o-NQDM is small, one might expect mixing of the singlet and triplet states, so the triplet state can be also accessible theoretically.
The transient absorption atλmax= 520 nm is assigned to
1o-NQDM 6 in agreement with the assignment made by Gisin
and Wirz[38]based on their observation of a diene in a ma-trix after photolysis of phthalazine. o-NQDM forms within 10 ns, which corresponds to decay of the S1 state. Similar
to the results of o-BQDM, the short-lived component of this transient is assigned to the singlet intermediate with a time constant 1.6s. The calculated wavelength for the vertical transition of singlet o-NQDM lies at 608 nm, and this value deviates from the observed by∼13%. This deviation is rela-tively large compared with the others shown in this work.
We conclude that production of singlet o-quinodimethanes (o-QDM) is observed from benzosultine and naphthosultine on UV excitation. Appearance of both o-QDM from these
sultines is around microsecond range. The results of the present work show that sultines serve as superior precursors for producing singlet o-quinodimethane, although one would consider sulfur to increase the rate of intersystem crossing so that triplet products become favored. No triplet o-QDM is produced from benzosultine. Isomerization of sultines to sulfolenes is expected to be inefficient due to a large bar-rier. The results of calculation show that the o-QDM has a singlet ground state and the triplet state has a biradicaloid structure. Optimized geometries and relative energies at level B3LYP/6-311++G** are obtained for dimers and spiro form of1o-BQDM.
Acknowledgments
We thank the National Science Council, Ministry of Ed-ucation (MOE program for promoting academic excellence of universities No.: 89-FA04-AA) of Republic of China and China Petroleum Company for financial support, and Na-tional Center of High-Performance Computing for support of computing facilities.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the online version, at 10.1016/j.jphotochem. 2004.08.004.
References
[1] J.L. Segura, N. Mart´ın, Chem. Rev. 99 (1999) 3199, and references therein.
[2] M.P. Cava, D.R. Napier, J. Am. Chem. Soc. 79 (1957) 1701. [3] M.P. Cava, D.R. Napier, J. Am. Chem. Soc. 80 (1958) 2255. [4] M.P. Cava, A.A. Deana, K. Muth, J. Am. Chem. Soc. 81 (1959)
6458.
[5] C.R. Flynn, J. Michl, J. Am. Chem. Soc. 95 (1973) 5802. [6] C.R. Flynn, J. Michl, J. Am. Chem. Soc. 96 (1974) 3280. [7] (a) V. Wintgens, J.C. Netto-Ferreira, H.L. Casal, J.C. Scaiano, J.
Am. Chem. Soc. 112 (1999) 2363;
(b) J.J. McCullough, Acc. Chem. Res. 13 (1980) 270;
(c) M. Fujiwara, K. Mishima, T. Tamai, Y. Tanimoto, K. Mizuno, Y. Ishii, J. Phys. Chem. A 101 (1997) 4912.
[8] J.L. Charlton, T. Durst, Tetrahedron Lett. 25 (1973) 5287. [9] F. Jung, M. Molin, R. Van der Elzen, T. Durst, J. Am. Chem. Soc.
96 (1974) 935.
[10] T. Durst, J.C. Huang, N.K. Sharama, D.J.H. Smith, Can. J. Chem. 56 (1978) 512.
[11] T. Durst, J.L. Charlton, D. Mount, Can. J. Chem. 64 (1986) 246. [12] (a) W.-S. Chung, W.-J. Lin, W.-D. Liu, L.-G. Chen, J. Chem. Soc.
Chem. Commun. (1995) 2537;
(b) W.-S. Chung, J.-H. Liu, Chem. Commun. (1997) 205; (c) J.-H. Liu, A.-T. Wu, M.-H. Huang, C.-W. Wu, W.-S. Chung, J. Org. Chem. 65 (2000) 3395;
(d) A.-T. Wu, W.-D. Liu, W.-S. Chung, J. Chin. Chem. Soc. 49 (2002) 77;
(e) W.-D. Liu, C.-C. Chi, I.-F. Pai, A.-T. Wu, W.-S. Chung, J. Org. Chem. 67 (2002) 9267.
[13] (a) P. Hohenberg, W. Kohn, Phys. Rev. 136 (1964) B864; (b) W. Kohn, L.J. Sham, Phys. Rev. 140 (1965) A1133.
[14] D.R. Salahub, M.C. Zerner (Eds.), The Challenge of d and f Elec-trons, ACS, Washington, DC, 1989.
[15] R.G. Parr, W. Yang, Density-Functional Theory of Atoms and Molecules, Oxford University Press, Oxford, 1989.
[16] R.E. Stratmann, J.C. Burant, G.E. Scuseria, M.J. Frisch, J. Chem. Phys. 106 (1997) 10175.
[17] R. Bauernschmitt, R. Ahlrichs, Chem. Phys. Lett. 256 (1996) 454. [18] M.E. Casida, C. Jamorski, K.C. Casida, D.R. Salahub, J. Chem.
Phys. 108 (1998) 4439.
[19] R.E. Stratmann, G.E. Scuseria, J. Chem. Phys. 109 (1998) 8218.
[20] M.D. Hoey, D.C. Dittmer, J. Org. Chem. 56 (1991) 1946. [21] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785. [22] A.D. Becke, Phys. Rev. A 38 (1988) 3098.
[23] B. Miehlich, A. Savin, H. Stoll, H. Preuss, Chem. Phys. Lett. 157 (1989) 200.
[24] A.D. Becke, J. Chem. Phys. 104 (1996) 1040. [25] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.
[26] P. Weber, J.R. Reimers, J. Phys. Chem. A 103 (1999) 9830. [27] J.P. Perdew, Phys. Rev. B 33 (1986) 8822.
[28] K. Burke, J.P. Perdew, Y. Wang, in: J.F. Dobson, G. Vignale, M.P. Das (Eds.), Electronic Density Functional Theory: Recent Progress and New Directions, Plenum, 1998.
[29] J.P. Perdew, in: P. Ziesche, H. Eschrig (Eds.), Electronic Structure of Solids, Akademie Verlag, Berlin, 1991, p. 11.
[30] J.P. Perdew, J.A. Chevary, S.H. Vosko, K.A. Jackson, M.R. Pederson, D.J. Singh, C. Fiolhais, Phys. Rev. B 48 (1993) 4978.
[31] J.P. Perdew, K. Burke, Y. Wang, Phys. Rev. B 54 (1996) 16533. [32] M.J. Frisch, G.W. Trucks, H.B. Schlegel, et al., Gaussian 98,
Revi-sion A.7, Gaussian, Inc., Pittsburgh, PA, 1998.
[33] (a) N.J. Turro, Modern Molecular Photochemistry, University Sci-ence, California, 1991 (Chapter 9);
(b) N.J. Turro, Modern Molecular Photochemistry, University Sci-ence, California, 1991, pp. 432–452 (Chapter 11);
(c) A.-T. Wu, W.-D. Liu, W.-S. Chung, J. Chin. Chem. Soc. 49 (2002) 77.
[34] S. Sakai, J. Phys. Chem. A 104 (2000) 11615. [35] L.A. Errede, J. Am. Chem. Soc. 83 (1961) 949.
[36] (a) W.R. Roth, M. Biermann, H. Dekker, R. Jochems, C. Mosselman, H. Hermann, Chem. Ber. 111 (1978) 3892;
(b) W.R. Roth, B.P. Scholz, Chem. Ber. 114 (1981) 3742. [37] N. Munzel, A. Schweig, Angew. Chem. Int. Ed. Engl. 26 (1987)
471.
[38] M. Gisin, J. Wirz, Helv. Chim. Acta 59 (1976) 2273.
[39] A. Ouchi, M. Sakurage, H. Kitahara, M. Zandomeneghi, J. Org. Chem. 65 (2000) 2350.
[40] A. Ouchi, Z. Li, M. Sakurage, T. Majima, J. Am. Chem. Soc. 125 (2003) 1104.
