Abstract

Pyrenes substituted with tert-butylethynyl, trimethylsilylethynyl, and trimethylsilylbutadiynyl groups were prepared, and the fluorescence yields (), lifetimes, and triplet-triplet absorption were measured in cyclohexane. Upon introduction of the groups possessing triple bond(s) to the pyrene skeleton, the fluorescence rate () increased. The variation of the terminating groups did not appreciably affect the and values. Increasing the number of the triple bond, the values increased by the magnitude of order whereas the were not varied. The effect of the ethynyl groups on the values was rationalized by the Strickler-Berg equation considering an increase of the 1 transition moment. Triplet-triplet absorption spectra of pyrene derivatives were obtained. The intersystem crossing rates () increased upon increasing the number of triple bonds terminated with the trimethylsilyl group whereas those between ethynylpyrenes were independent of the terminating groups. Heavy atom effect failed to rationalize the enhancement of the values upon adding the triple bond to the pyrene moiety.

1. Introduction

Recently, there have been a number of synthetic reports on silyl-substituted aromatic compounds that could be candidates for high-yield fluorescent materials [117]. The increase in the fluorescence yield was explained by two ways [18]. One is that introduction of silyl groups into condensed aromatic systems causes a bathochromic shift, which increases the molar absorption coefficients (ε) of the resultant systems. Substitution with silyl groups, such as trialkylsilyl and triphenylsilyl groups, considerably increases the fluorescence rates or quantum yields because the fluorescence rate constants () are linearly related to ε by the Strickler-Berg equation [19]. The other is from the viewpoint of deactivation profiles in excited states. It is generally considered that deactivation from the lowest excited singlet states (S1) of aromatic compounds in solution proceeds via three competitive photophysical pathways, fluorescence, internal conversion (IC) to the ground state (S0), and intersystem crossing (ISC) to the lowest triplet state (T1) [20, 21]. Silyl substitution of aromatic compounds stabilizes the S1state, which is lower in energy than the second triplet (T2) state. Therefore, the ISC rate is negligibly small compared to the rates of other processes. Some aromatic compounds have negligible rates of IC from the S1 to the S0 states. In these cases, a considerable increase in the fluorescence yield is expected [18].

In this decade, silylethynyl groups have been introduced to fundamental aromatic skeletons, such as anthracene [2224], pentacene [7], and pyrene [15]. These substituted compounds show intense fluorescence compared to the parent aromatic compounds. The mechanism for an increase of the fluorescence yield in these compounds has not been revealed in detail. To understand the effect of ethynyl groups on excited aromatic compounds, systematic investigations are necessary based on the photophysical parameters of the deactivation processes from the excited states.

In the present work, we prepared a series of pyrenes with trimethylsilylethynyl (TMSE), tert-buthyletynyl (BuE), and trimethylsilylbutadiynyl (TMSB) groups (Scheme 1). The fluorescence spectra, yields, lifetimes, and triplet-triplet absorption spectra of these compounds were measured in solution. Based on the photophysical parameters of the rates of fluorescence and nonradiative processes, the effects of introducing these substituent groups on the fluorescent states were studied. In the text, character for the transition moments in the excited singlet states of the pyrene derivatives used in the present work is expressed as and with the directions as shown in Scheme 1.

103817.sch.001

2. Experimental Section

2.1. Instruments

Absorption spectra were recorded on a JASCO Ubest 50 spectrophotometer. Emission spectra were recorded on a Hitachi Fluorescence Spectrometer F-7000. Nanosecond fluorescence lifetimes () were determined using a time-correlated single-photon counting fluorimeter (FL-900CDT, Edinburgh Analytical Instruments). Picosecond fluorescence lifetimes were determined with a femtosecond laser system equipped with a mode-locked Ti:sapphire laser (Tsunami, Spectra Physics), which had a center wavelength of 800 nm, pulse width of 70 fs, and repetition rate of 82 MHz, with a CW green laser (532 nm, 4.5 W, Millenia V, Spectra Physics) for pumping. The excitation and detection systems for obtaining the time profile of fluorescence are detailed elsewhere [25]. The fluorescence lifetime and transient absorption measurements were carried out at 295 K. The fluorescence quantum yields () were determined by using an absolute luminescence quantum yield measurement system (C9920-02) from Hamamatsu Photonics [26]. The fourth harmonic (266 nm) of an Nd3+:YAG laser (JK Lasers HY-500; pulse width 8 ns) was used as the light sources for flash photolysis. The number of laser pulses in each sample was less than four to avoid excess exposure. The details of the detection system for the time profiles of the transient absorption have been reported elsewhere [27]. The transient absorption spectra were obtained using a USP-554 system from Unisoku, which provides a transient absorption spectrum with one laser pulse. Melting points were determined on a Yanagimoto micro melting point apparatus (Yanaco MP-500). 1H and 13C NMR spectra were recorded on a Varian MERCURY-300 (300 MHz and 75 MHz, resp.) spectrometer using Me4Si as an internal standard. IR spectra were determined on a Jasco FT/IR-230 spectrometer. Mass spectra (EI) were taken on a SHIMADZU GCMS-QP5050 operating in the electron impact mode (70 eV) equipped with GC-17A and DB-5MS column (J&W Scientific Inc., Serial: 8696181). High-performance liquid chromatography (HPLC) separations were performed on a recycling preparative HPLC equipped with Jasco PU-2086 Plus pump, RI-2031 Plus differential refractometer, and Megapak GEL 201F columns (GPC) using CHCl3 as an eluent.

2.2. Materials

Cyclohexane (CH, spectroscopic grade from Sigma-Aldrich), methylcyclohexane (Uvasol from Dojin), and isopentane (UV-spectroscopy grade from Fluka) were used as supplied. CH was used as the solvent at 295 K whereas a mixture of methylcyclohexane and isopentane (3 : 1 v/v) was used as a matrix for the phosphorescence measurement at 77 K. Pyrene was purchased from commercial source and used without further purification. 1-(Trimethylsilylethynyl)pyrene (TMSEPy) was synthesized according to the published procedure [15]. 1-(3,3-Dimethylbut-1-ynyl)pyrene (BuEPy) and 1-(trimethylsilylbut-1,3-diynyl)pyrene (TMSBPy) were prepared by procedures described below. All the samples for fluorescence and transient absorption measurements were degassed in a quartz cell with a 1 cm path length by several freeze-pump-thaw cycles on a high vacuum line. The concentrations of the samples for emission measurements were in the magnitude of 10−6 mol dm−3 to avoid formation of excimer fluorescence while those for laser photolysis were adjusted to achieve optical density of 1.0 at the excitation wavelength (266 nm).

1-(3,3-Dimethylbut-1-ynyl)pyrene (BuEPy)
To a round-bottom flask equipped with a magnetic stirrer, pyrene (5.0 g, 25.0 mmol), N-bromosuccinimide (3.6 g, 20 mmol) and DMF (100 mL) were placed. The solution was stirred for 17 h at room temperature. CHCl3 and H2O were added and the mixture was shaken. The organic layer was separated, dried over Na2SO4, filtered, and evaporated. Purification by column chromatography on silica gel (eluent; hexane/EtOAc = 3/1, = 0.93) followed by recrystallization from EtOH gave 1-bromopyrene as a yellow solid (1.3 g, 23% yield). Data for 1-bromopyrene: 1H NMR (300 MHz, CDCl3) δ 7.98–8.11 (m, 4H), 8.17–8.25 (m, 4H), and 8.44 (d, J = 9.2 Hz, 1H) ppm. To an argon-purged, round-bottom flask equipped with a Dimroth condenser, a part of the above 1-bromopyrene (380 mg, 1.4 mmol), 3,3-dimethyl-1-butyne (1.5 mL, 12 mmol), CuI (138 mg, 0.72 mmol), Pd(PPh3)2Cl2 (56 mg, 0.08 mmol) and piperidine (15 mL) were placed. The solution was heated and vigorously stirred for 14 h under reflux. Evaporation of the mixture gave a residue, which was subjected to column chromatography on silica gel (eluent: CHCl3/hexane = 1/4, = 0.81). The product was further purified by HPLC (GPC) separation followed by column chromatography on silica gel (eluent: hexane) to give 1-(3,3-dimethylbut-1-ynyl)pyrene (BuEPy, 227 mg, 60% yield). Data for BuEPy: colorless solid; mp 105-106°C; 1H NMR (300 MHz, CDCl3) δ 1.49 (s, 9H), 7.95–8.00 (m, 3H), 8.04–8.05 (m, 2H), 8.10–8.18 (m, 3H), 8.52 (d, J = 9.1 Hz, 1H) ppm; 13C NMR (75 MHz, CDCl3) δ 28.87, 31.68, 104.59, 118.84, 114.37, 124.47, 125.21, 125.26, 125.62, 126.02, 127.23, 127.58, 127.90, 129.49, 130.57, 131.07, 131.23, 131.69 ppm; GC-MS (EI) m/z (relative intensity, %) 267 (100), 282 (83), 252 (35); IR (KBr) ν 2227 (CC) cm−1.

1-(Trimethylsilylbut-1,3-diynyl)pyrene (TMSBPy)
To a round-bottom flask equipped with a magnetic stirrer, (trimethylsilylethynyl)pyrene (TMSEPy, 320 mg, 1.1 mmol) [15], 1 M tetrabutylammonium fluoride (TBAF) solution in THF (1.5 mL, 1.5 mmol), and additional THF (10 mL) were placed and the solution was stirred for 4 h at room temperature. The solvent was removed by evaporation. Purification by column chromatography on silica gel (eluent: CHCl3, = 0.96) gave crude 1-ethynylpyrene (242 mg). To a round-bottom flask equipped with a magnetic stirrer, the crude 1-ethynylpyrene (242 mg), CuI (190 mg, 1.9 mmol), trimethylsilylacetylene (0.5 mL, 3.5 mmol), TMEDA (2 mL), and CH2Cl2 (20 mL) were placed. The mixture was stirred for 4 h at room temperature. The mixture was filtered using celite. Purification by column chromatography on silica gel (eluent: CHCl3, = 0.77) followed by HPLC (GPC) and recrystallization from EtOH gave 1-(trimethylsilylbut-1,3-diynyl)pyrene (TMSBPy, 67 mg, 19%). Data for TMSBPy: orange solid; mp 137-138°C; 1H NMR (300 MHz, CDCl3) δ 0.31 (s, 9H), 8.01–8.25 (m, 8H), 8.55 (d, J = 9.3 Hz, 1H) ppm; 13C NMR (75 MHz, CDCl3) δ 0.14, 79.65, 88.28, 92.32, 115.60, 124.06, 124.28, 124.45, 125.26, 125.88, 125.96, 126.38, 127.12, 128.79, 128.85, 130.68, 130.92, 131.89, 133.52 ppm; GC-MS (EI) m/z (relative intensity, %) 322 (100), 307 (84); IR (KBr) ν 3041, 2955, 2191, 2090, 1595, 1274, 1248, 1070, 1018, 964, 838 cm−1.

3. Results and Discussion

Figure 1 shows absorption, fluorescence, and phosphorescence spectra of pyrene and the substituted ones with BuE, TMSE, and TMSB groups.

All the absorption and emission spectra showed well-resolved vibrational structures. The measured excitation spectra for the fluorescence and phosphorescence were found to agree with the corresponding absorption spectra. Excimeric fluorescence was avoided due to the low concentrations (~10−6 mol dm−3). The 0-0 origins of the absorption and fluorescence spectra were clearly seen except for the absorption of BuEPy, whose band may be overlapped by the band. Accordingly, the S1S0 transition of the pyrenes used in the present work can be regarded as that of the transition. The fluorescence spectra were mirror imaged by the absorption band. The S1 state energies () of the compounds were determined from the averaged energies of the 0-0 origins of the corresponding absorption and fluorescence spectra. The triplet energies () of pyrene and BuEPy were determined from the origins of the phosphorescence spectra. Phosphorescence from TMSEPy and TMSBPy was not observed. The obtained and values are listed in Table 1. The obtained values seemed to decrease on being substituted with triple bond(s).

The quantum yields () of nonradiative deactivations of the studied compounds were estimated by using (1) and fluorescence yields () Based on these quantum yields, the corresponding rates ( and ) were, respectively, estimated with (2) and (3) using the fluorescence lifetimes (). By introduction of triple bonds to the pyrene moiety, a small increase in fluorescence yields was seen in TMSEPy and TMSBPy (0.70) compared to the fluorescence yield (0.64) of pyrene whereas a decrease of the fluorescence yield was found for BuEPy (0.45). Also only a small difference in the and values (~106 s−1) between BuEPy and TMSEPy was seen, indicating that the terminating group (tert-butyl and TMS) has little influence on these rates. On increasing the number of the triple bond in the substituent group, the and values increased from 106 s−1 to 107 or 108 s−1. The values are closely related to the direction and the magnitude of the transition moment [20]. The increase in by the introduction of TMSE, BuE, and TMSB groups onto the pyrene skeleton can be interpreted as an increase in the transition moment on extending π-electron systems by substitution with triple bonds. The absorption spectra of pyrenes used in this work showed that the S1S0 character was due to the transition. The small change in the values between pyrene and ethynyl pyrenes may derive from the unfavorable substituted position for enhancing the transition moment. Even at the 1-position, the degree of the effect of the butadiynyl group, which looks like a tandem diethynyl, on the fluorescence pathway may be different from that of the single ethynyl group.

The quantum yield () of ISC of pyrene in CH is reported to be [28], which is coincident with the value of (0.36) determined in the present work. Therefore, (4) is obtained for pyrene, Equation (4) indicates that the IC rate () from the S1 state of pyrene to the ground state can be negligible compared to those of the fluorescence and ISC processes (Ermolaev’s rule [21]). In the present work, the ISC processes of the pyrene derivatives were investigated by observing the triplet-triplet (T-T) absorption using flash photolysis techniques. Figure 2 shows the transient absorption spectra obtained at 500 ns upon laser photolysis of CH solutions of the pyrene derivatives.

Transient absorption spectra were obtained in the wavelength region of 430–560 nm. The intensities of all the transient absorption decreased with lifetimes () of a few tens of microseconds (see Table 1), and the decay was accelerated in the presence of the dissolved oxygen. From these observations, the obtained transient signals could be ascribed to the T-T absorption of the corresponding pyrene compounds. The wavelengths of the maximum absorption () for the T-T absorption spectra are listed in Table 1. It is noted that only a little difference in the shape of the T-T absorption spectrum between BuEPy and TMSEPy can be seen. This observation indicates that the terminating group (tert-butyl and TMS) has little influence on electronic character of the triplet state of the pyrene moiety through the triple bond. The intensity of the triplet absorption of the studied pyrenes was found not to vary on increasing the number of the C-C triple bonds. Assuming that the molar absorption coefficients of triplet pyrenes with triple bond(s) are similar to those of triplet pyrene, it is indicative from the observation of the T-T absorption that the values of the pyrene derivatives may be almost the same as those of pyrene. This evaluation agrees with the trend in the values determined from the values by (1). Plausibly, Ermolaev’s rule can be also applicable to the pyrene derivatives substituted with the triple bond(s). This consideration is qualitatively indicative that introduction of the triple bond enhances the rate () of ISC.

In general, the ISC mechanism is interpreted in terms of the spin-orbit interaction, and the is expressed by Fermi’s golden rule [21], In (5), and are the wave functions for the initial excited singlet state (S1) and the final triplet state (T1), respectively; is the density of the final state; and are the Hamiltonian for the spin-orbit interaction and the Franck-Condon factor between the initial and final states, respectively. The perturbation Hamiltonian () can be expressed with the use of the angular- and spin-momentum operators ( and , resp.) and the spin-orbit coupling constant (), which gives a measure for the heavy atom effect, Equation (5) can then be rewritten as follows: Here, the values for C, H, and Si atoms are known to be 32, 0.24, and 130 cm−1, respectively [29]. The sums () of the squared values for all the compound used in the present work are listed in Table 1. The ratios of for BuEPy, TMSEPy, and TMSBPy against pyrene were estimated to be 1.4, 2.3, and 2.5, respectively while those of the values were 4.7, 4.2, and 53, respectively. The estimated ratios for do not agree with those for (). It is inferred from these considerations that the mechanism of ISC enhancement by the introduction of triple bond(s) cannot be simply interpreted by Fermi’s golden rule. Presumably, other interactions caused by the introduction of the triple bonds, such as vibronic coupling of the triple bonds to the pyrene moiety, would be the origins for the increment in the ISC processes. However, in the present study, we were unable to proceed further investigations of the effect of C-C triple bonds in solution. Multiethynyl- and hexatriynylpyrenes would be suitable compounds to understand the effect on the photophysical processes of pyrene, which is in progress.

4. Conclusions

We prepared pyrenes having C-C triple bonds and investigated their deactivation properties of the fluorescent state based on the measurements of the fluorescence yields () and lifetimes () and triplet absorption. The absorption and fluorescence features of these derivatives were the same as those of pyrene. Although the values were not drastically varied with an increase of the number of the triple bonds, the value increased. The variation of the terminating groups on the ethynyl moiety did not appreciably affect the rates of fluorescence and nonradiative processes. The enhancement by the introduction of triple bonds was explained in terms of an increase of the transition moment in the pyrene moiety. Based on the measurements of the transient absorption, it was indicative that Ermolaev’s rule is applicable to the pyrene derivatives studied in this work. The value was found to increase as an increase of the number of the triple bonds. The heavy atom effect was not sufficient to explain the trend of this enhancement.

Acknowledgments

This work was supported by Grants-in-Aid for Scientific Research (KAKENHI) from JSPS (no. 23350059) and the “Element Innovation” Project by the Ministry of Education, Culture, Sports, Science, and Technology, Japan (to M. Yamaji). H. Maeda and K. Mizuno acknowledge Grant-in-Aids for Scientific Research (KAKENHI) from JSPS (no. 20550049 and 23550047), and Cooperation for Innovative Technology and Advanced Research in Evolutional Areas (CITY AREA) program from the Ministry of Education, Culture, Sports, Science, and Technology, Japan. H. Maeda also thanks the Mazda Foundation for the research grant. M. Yamaji thanks Dr. Atsushi Kobayashi at Gunma University for his assistance on determining the fluorescence yields with the instrument (C9920-02).