PCCP
Photoinduced electron transfer between metal octaethylporphyrins and fullerenes (C60/C70) studied by laser flash photolysis: electron-mediating and hole-shifting cycles Mohamed E. El-Khouly, Yasuyuki Araki, Mamoru Fujitsuka and Osamu Ito* Institute of Multidisciplinary Research for Advanced Materials, Tohoku University, CREST, Japan Science and Technology, Katahira, Aoba-ku, Sendai 980–8577, Japan. E-mail:
[email protected] Received 22nd February 2002, Accepted 7th May 2002 First published as an Advance Article on the web 5th June 2002
Electron transfer (ET) processes of fullerenes (C60/C70) with metal octaethylporphyrins (MOEP) in a polar solvent have been investigated by a nanosecond laser photolysis technique in the visible/near-IR regions. By the selective excitation of C60/C70 using OPO laser light it has been proved that electron transfer takes place from the ground states MOEP to the triplet excited states 3C60*/3C70*. By selective excitation of MOEP the electron transfer processes via 3MOEP* to C60/C70 were also confirmed. In nonpolar toluene the lifetimes of 3 C60*/3C70* decreased in the presence of MOEP without evidence for electron and/or energy transfer processes. On adding benzonitrile to toluene the ion-radical formation was increased, suggesting that the triplet exciplexes 3 [MOEP(d+)—C60(d)/C70(d)]* would be dominant prior to the ion-pair formation in a less-polar solvent. On addition of a viologen dication the electron of the anion radicals of C60/C70 transfers to the viologen dication yielding the viologen radical cation. In the system MOEP–C60/C70–aromatic amine (DTQH), as a second donor having lower oxidation potential than MOEP, the hole shift from the radical cation of MOEP to the aromatic amine was observed. These observations proved that the photosensitized electron-transfer/electron-mediating and photosensitized electron-transfer/hole-shift cycles have been confirmed by the transient absorption spectral method.
Introduction Electron transfer between a redox pair has long been of interest in chemistry and biology because it is a universal and fundamental phenomenon in nature.1,2 Because fullerenes and porphyrins have unique photophysical and photochemical properties, they have been utilized as important compounds for the design of novel artificial photosynthetic systems.3–11 Since light absorption by porphyrins and fullerenes covers a broad range of the visible wavelength regions and intersystem crossing (isc) occurs with close to 100% quantum yield, these systems have been the subjects of extensive work to achieve highly efficient photo-energy conversion systems. Furthermore, fullerenes such as C60 and C70 are known to enhance the photoinduced electric conductivity when they are added to phthalocyanines and porphyrins.12–15 In our continuous efforts to understand the mechanism of ET processes of fullerene–porphyrin systems, we have reported the ET processes of mixtures of fullerenes (C60/C70) with phthalocyanines16,17 and porphyrins18,19 via their triplet states in polar solvents. We only compared the ET ability of central zinc metal derivatives with free-metal derivatives in the ground and triplet excited states. In the case of chlorophylls,20 which contain central Mg metal, similarly high electron donor ability to that of central Zn metal has been pointed out. In the present study we shed light on the effect of central metal ions in the electron transfer processes of fullerenes (C60/C70) and metal octaethylporphyrins MOEP (where, M ¼ PdII, MgII, ZnII, NiII, CoII, CuII and V=O) via 3 C70*/3C60* and 3MOEP*. PdOEP, MgOEP and ZnOEP are diamagnetic, while NiOEP, CoOEP, CuOEP and (V=O)OEP are paramagnetic (Fig. 1). 3322
In addition, we found that the excitation of either porphyrins or C60/C70 in nonpolar solvents does not provide any evidence for energy transfer and/or electron transfer processes, although the quenching rates of the triplet states were accelerated. In the present study, we will try to disclose this phenomenon using a laser flash photolysis technique, assuming that the existence of exciplex intermediate accelerates deactivation of the triplet states in nonpolar solvent.
Fig. 1
Phys. Chem. Chem. Phys., 2002, 4, 3322–3329 This journal is # The Owner Societies 2002
Investigated molecules.
DOI: 10.1039/b201922c
The electron mediating process from C60 /C70 to octylviologen dication (OV2+) was also investigated to prove a photosensitized electron-transfer/electron-mediating cycle of the C60/C70–MOEP–OV2+ system. Furthermore, the holeshift process has been examined on addition of aromatic amine under photoexcitation of MOEP in the presence of C60/C70 to prove the efficient photosensitizing electron-transfer/hole-shift cycle.
Experimental Materials C60 and C70 (>99%) were purchased from Texas Fullerene Corp. Metal octaethylporphyrin (MOEP, Aldrich), benzonitrile (BN) (99.9% HPLC grade, Aldrich) and toluene (99.8% HPLC grade, Aldrich) were used as received. As a hole-shift reagent, we employed N,N-diphenyl-N-(1,2,3,4-tetrahydroquinolin-6-yl-methylene)-hydrazine (DTHQ) (Fig. 1), which was commercially available (Anan Koryo LTD, Japan). Octylviologen dication (OV2+) with perchlorate anion as a counter ion was employed as an electron-mediating reagent, and was prepared from commercially available methyl viologen dication with chloride anions. Methods Steady-state absorption spectra were measured using an optical cell (0.2–1.0 cm) with a JASCO V-570 spectrophotometer. Cyclic voltammetry (CV) measurements were performed with a BAS CV-50W Voltammetric Analyzer (Japan). All measurements were carried out under argon in BN solution containing 0.1 103 mol dm3 tetra-n-butylammonium perchlorate (Bu4NClO4) as the supporting electrolyte and an AgCl reference electrode. Transient absorption spectra and time profiles were measured using the light from an OPO laser (Continuum, fwhm 6 ns and 21 mJ pulse1) as an excitation source. For time-scale measurements longer than 10 ms, InGaAs-PIN and Si-PIN photodiodes were employed as detectors in the near-IR and visible regions, respectively, to monitor the steady light from a continuous Xe-lamp (150 W). More details of the experimental set-up have been given elsewhere.21 The measurements were carried out at 23 C using freshly prepared argon-saturated solutions to eliminate the influence of the O2 effect.
in BN. The absorption spectrum of the mixture of C60/C70 and MOEP is the same as the calculated spectrum of the corresponding components, suggesting that the interaction between C60/C70 and MOEP in the ground state is negligibly weak. In the C70–MOEP system, C70 was predominantly excited by applying 470 nm laser light due to the higher molar absorption coefficient compared to MOEP. On the other hand, MOEP compounds were predominantly excited by laser light tuned at the absorption maximum in the 540–580 nm region in the presence of C60 and C70 . For C60–MOEP systems, although the molar absorption coefficient of C60 is quite low compared to MOEP in the visible region, it is also possible to measure the electron transfer process via 3C60* for some MOEP (M ¼ CoII, NiII, CuII), since these metalloporphyrins do not show any transient absorption bands on the nanosecond time-scale. Thus, the absorption bands of 3C60*, C60 and MOEP + were observed clearly. Electron transfer from MOEP to 3C70* in a polar medium By the use of a nanosecond laser pulse, the electron transfer process via the triplet states can be observed in real time. Such measurements have made it possible to understand the details of the ET processes occurring in solutions. By photoexcitation of C70 (0.1 103 mol dm3) in deaerated BN using 470 nm laser light (Fig. 3(a)), the transient absorption spectrum immediately after the laser pulse exhibited only an absorption band at 980 nm which is unambiguously assigned to the triplet state 3 C70*.4,9,22,23 On substituting the reported extinction coefficient (e ¼ 6500 mol1 dm3 cm1)16–20,22–24 into the observed initial absorbance at 980 nm, the initial concentration of 3 C70* generated by one laser exposure was calculated as 1.8 105 mol dm3, which implies that 18% of C70 are photoexcited to 3C70* by one laser shot with 5 mJ power. By photoexcitation of C70 (0.1 103 mol dm3) in the presence of MgOEP (0.1 103 mol dm3) in BN by applying 470 nm (Fig. 3(b)), the decay rates of 3C70* at 980 nm were significantly accelerated. With the decay of 3C70*, the concomitant rises of C70 at 1380 nm16–20,22–24 and MgOEP + at 650 nm
Results and discussion Steady-state absorption spectra Fig. 2 shows the absorption spectra of the same concentrations (0.05 103 mol dm3) of C60 , C70 , PdOEP and (V=O)OEP
Fig. 2 Absorption spectra of C60 , C70 , PdOEP and (V=O)OEP in BN solution. The employed concentration is 0.05 103 mol dm3.
Fig. 3 Transient absorption spectra observed by 470 nm laser excitation of (a) C70 (0.1 103 mol dm3) and (b) C70 (0.1 103 mol dm3) in the presence of MgOEP (0.1 103 mol dm3) in Arsaturated BN solution. Inset: Time profiles at 650, 980 and 1380 nm.
Phys. Chem. Chem. Phys., 2002, 4, 3322–3329
3323
absorption spectra of C70 in the presence of PdOEP, ZnOEP, (V=O)OEP, CoOEP, NiOEP and CuOEP displayed the same features as in the MgOEP case. A more detailed picture of the kinetic event is shown in Fig. 4, where the rate constants of the bimolecular quenching (kq) were evaluated under the pseudo-first-order condition [3C70*] [MgOEP] by monitoring the decay of 3C70* as a function of [MgOEP]. The decay time profiles of 3C70* obey first-order kinetics; each rate constant is referred to (k1st). The linear concentration-dependence of the observed k1st values gives the kq values, as listed in Table 1. The efficiency of ET can be calculated from the ratio of the maximal concentration of the generated radical ions to the initial concentration of 3C70*. The ratios of [C70 ]max/ [3C70*]max are plotted against the concentration of MOEP as shown in Fig. 5. The saturated values are attributed to the Fet values (in Table 1). Furthermore, the (ket) values were finally obtained from the following equation, ket ¼ kq Fet . The evaluated ket and Fet values are summarized in Table 1, in which the ket values (4.3 108–1.5 109 mol1 dm3 s1) are only slightly smaller than the diffusion-controlled limit (kdiff ¼ 5.6 109 mol1 dm3 s1) in BN.26 Table 1 shows clearly that the observed Fet values are less than unity, suggesting that there are some deactivation routes for 3C70* other than by electron-transfer. In general, collisional quenching and/or an encounter complex can be considered as deactivation processes. The possibility of the energy transfer process from 3C70* to MOEP in BN solution is quite low owing to the absence of the rise of 3MOEP* at 440 and 760 nm.17,18 Thus, the deactivation process may be attributed to the formation of an encounter complex or collisional quenching like the exciplex 3[C70(d)—MOEP(d+)]*. The forthcoming sections discuss this observation on changing the solvent polarity. The Fet values via 3C70* varied with the central metal according to the following order; PdOEP > MgOEP > ZnOEP > (V=O)OEP > CoOEP > NiOEP > CuOEP, as listed in Table 1. The change in the donor abilities of MOEP may be mainly explained by their Eox values. From Table 1, it is seen that the ket and Fet values gradually increase with decreasing E(D/D +), except for (V=O)OEP, indicating that the ET process from MOEP to 3C70* might be the mechanism responsible for the quenching in the present systems. The feasibility of the ET process from the ground state MOEP to 3C70* is directed by the standard Gibbs energy change (DGet0), which can be expressed by the Rehm–Weller equation:
Scheme 1
Fig. 4 Dependence of time profiles at 980 nm on MgOEP concentrations (0.0–0.2 103 mol dm3) in the 3C70*–MgOEP system in Arsaturated BN. Inset: Pseudo-first-order plot.
were observed, as shown in the inserted time profiles in Fig. 3(b). The decay of 3C70* and the rises of C70 and MgOEP + seem to match each other. These observations indicate that the ET takes place via 3C70*, as shown in Scheme 1. Furthermore, the contribution of 3C70* to the ET process was confirmed by the O2-effect on the yields of C70 and MgOEP +. The decay of 3C70* was accelerated on addition of O2 in the absence of MgOEP, indicating that 3C70* was quenched owing to energy transfer to O2 , yielding the singlet oxygen (1O2);25 kO2 was evaluated to be 8.2 109 mol1 dm3 s1 assuming [O2] ¼ 1 103 mol dm3 in BN.26 Addition of O2 under the condition of [O2] > [MgOEP] obviously suppressed the formation of C70 and MgOEP +, supporting the ET process via 3C70* as shown in Scheme 1. The transient
DG et 0 ðkcal mol1 Þ ¼ 23:06½E ox ðD=D þ Þ
E red ðA =AÞ e2 =erDA E T
ð1Þ
Table 1 Oxidation potential of MOEP (Eox), Gibbs energy changes (DGet0), triplet quenching rate constants (kq), electron-transfer quantum yields (Fet), electron-transfer rate constants (ket) for 3C60*/3C70* with MOEP and back electron-transfer rate constants (kbet) for C60 /C70 with MOEP + in Ar-saturated BN System
Eox/V
DGet0/kcal mol1
kq/mol1 dm3 s1
Feta
ket/mol1 dm3 s1
kbet/e /mol1 dm3 s1
kbet/mol1 dm3 s1 a
3
0.44 0.53 0.63 0.64 0.68 0.85 0.96 0.68 0.77 0.85
15.9 13.8 11.5 11.3 10.4 6.5 4.1 8.5 6.5 4.5
2.2 109 2.4 109 2.9 109 2.7 109 2.2 109 2.0 109 1.8 109 3.3 109 3.5 109 2.6 109
0.74 0.52 0.40 0.32 0.39 0.21 0.40 0.11 0.11 0.06
1.6 109 1.2 109 1.1 109 8.7 108 8.5 108 4.3 108 7.2 108 3.6 108 3.9 108 1.6 108
8.0 105 1.2 106 2.2 106 1.6 106 1.0 106 2.0 106 1.2 106 6.5 105 1.0 105 8.0 105
3.2 109 4.7 109 9.0 109 6.5 109 4.0 109 8.0 109 4.6 109 7.8 109 1.2 109 9.7 109
C70*/PdOEP C70*/MgOEP 3 C70*/ZnOEP 3 C70*/NiOEP 3 C70*/CoOEP 3 C70*/CuOEP 3 C70*/(V=O)OEP 3 C60*/CoOEP 3 C60*/NiOEP 3 C60*/CuOEP 3
a These values were calculated employing the reported molar extinction coefficients; e(3C70*) ¼ 6500 mol1 dm3 cm1 at 980 nm, e(C70 ) ¼ 4000 mol1 dm3 cm1 at 1380 nm, e(3C60*) ¼ 16 100 mol1 dm3 cm1 at 740 nm, and e(C60 ) ¼ 12 000 mol1 dm3 cm1 at 1080 nm.4,7,9,16–23
3324
Phys. Chem. Chem. Phys., 2002, 4, 3322–3329
Fig. 5 Dependence of electron-transfer efficiency on concentration of MOEP for 3C70*–MOEP systems in Ar-saturated BN.
Where Eox(D/D +), Ered(A /A), ET and e2/erDA are the oxidation potential of the donors (MOEP), the reduction potential of the acceptor (C70), the triplet energy of the exciting species (3C70*) and the Coulombic term, respectively. The estimated DGet0 values are summarized in Table 1, employing ET(3C70*) ¼ 1.50 eV16–23 and Ered(A /A) ¼ 0.43 V vs. SCE, Coulomb energy ( ¼ 0.06 eV in BN)24,27 in addition to the Eox(D/D +) values.28 The observed ket values are plotted against DGet0 in Fig. 6. The ket values increase with decreasing DGet0 along the curve calculated by the semiempirical Rehm– Weller plot. The calculated DGet0 values for 3C70*–MOEP systems are all negative, from which it would be anticipate that both kq and ket values are close to kdiff . The observed kq values for 3C70*–MOEP systems are all in a similar region (2.0– 2.7 109 mol1 dm3 s1), despite considerable variations in the DGet0 values from 15.9 to 3.92 kcal mol1. Electron transfer via 3C60* to MOEP with paramagnetic central metal Although the steady-state absorption spectra show that the molar absorption coefficient values of NiOEP, CuOEP and CoOEP are much higher than C60 in the 530–550 nm region, no transient absorption bands due to 3NiOEP*,3CoOEP* and 3CuOEP* were observed in the 400–450 nm region. In the case of metalloporphyrins with unpaired d-electron(s), it was found that the excited electronic states become complicated due to the interaction of the unpaired d-electrons with the porphyrin p-electron system.29–31 Darwent et al.32 reported that porphyrins with paramagnetic transition metals have very short excited-state lifetimes, which essentially precludes their use as sensitizers in fluid solution at room temperatures.
Fig. 6 Plot of ket vs. DGet0 for the electron transfer of 3C70*–MOEP systems in Ar-saturated BN. The solid line was calculated from the ˚ ). Rehm–Weller equation assuming center–center distance (RCC ¼ 8 A
In this respect, the ET processes via 3C60* with NiOEP, CoOEP and CuOEP have been investigated by applying 550 nm laser-light excitation in BN. The transient spectrum shows the absorption maximum at 740 nm with a weak band at 550 nm that unambiguously is assigned to 3C60*,33–35 while no trace of absorption of 3MOEP* appeared (Fig. 7). With the decay of 3C60*, the absorption intensity increases at 1080 nm, which is ascribed to C60 .33–35 The time profiles in Fig. 7 support that there is an ET process from CoOEP to 3C60*. After the complete decay of 3C60*, the absorption bands remained at 600 nm and 750–800 nm in addition to 1080 nm; these bands in the visible region may be attributed to the presumed CoOEP +. Similar behavior was observed for NiOEP and CuOEP. Adding O2 to the solutions suppresses the formation of C60 , as well as increasing the quenching rate of 3C60* through energy transfer. These observations indicate that ET takes place from ground state MOEP (where M ¼ CoII, CuII and NiII) to 3C60*, as in Scheme 1. In the absence of O2 , the kq values are obtained from the concentration dependence of the decay rates of 3C60* in the presence of CoOEP, NiOEP and CuOEP under the pseudofirst condition [3C60*] [MOEP]. The Fet values for the ET process via 3C60* were also evaluated by plotting [C60 ]max/ [3C60*]max against [MOEP] using reported values for e(C60 ) and e(3C60*).4,8,16–23 The Fet and ket values in the 3C60*– MOEP (M ¼ CuII, NiII and CoII) systems are listed in Table 1. Compared with the corresponding 3C70*/MOEP systems, the Fet and ket values for 3C60*/MOEP systems are small, reflecting the less negative DGet0 values of the 3C60*– MOEP systems. Electron transfer via 3MOEP* with diamagnetic central metal to C60/C70 According to the steady-state absorption spectra given in Fig. 2, the molar absorption coefficients of MOEP are much higher than C60 and C70 over the region 550–600 nm leading to predominant excitation of MOEP (where M ¼ MgII, PdII, ZnII and V=O). Fig. 8 represents the transient absorption spectra after laser light irradiation of MgOEP in the presence of C60 in BN. The transient spectrum at 1 ms shows a maximum at 440 nm due to 3MgOEP*. In the spectrum at 10 ms, the presence of C60 was observed by build up of absorption at 1080 nm that parallels a concomitant decay of 3MgOEP*. The decay of 3MgOEP* at 440 nm was further accelerated on addition of oxygen to the solution, resulting in decrease in C60 . Considering these observations, it is obviously shown
Fig. 7 Transient absorption spectra observed by 530 nm laser excitation of C60 (0.1 103 mol dm3) in the presence of CoOEP (0.1 103 mol dm3) in Ar-saturated BN. Inset: Time profiles at 740 and 1080 nm.
Phys. Chem. Chem. Phys., 2002, 4, 3322–3329
3325
Scheme 3
Fig. 8 Transient absorption spectra observed by 550 nm laser excitation of MgOEP (0.1 103 mol dm3) in the presence of C60 (0.1 103 mol dm3) in Ar-saturated BN. Inset: Time profiles at 440 and 1080 nm.
that the electron transfer takes place from 3MOEP* to C60 as shown in Scheme 2. A possibility of quenching porphyrins singlet states is excluded due to the slow rise of C60 . These kinetic parameters for 3MOEP*–C60 systems are listed in Table 2. The kq values via 3MOEP* are almost the same as those via 3C70*/3C60*, which is reasonable on the basis of their similar DGet0 values. It is remarkable that the Fet values via 3MOEP*–C70 systems seem to be higher than for 3 MOEP*–C60 systems; the difference can be explained by the difference in the Ered values of C60 (0.51 V vs. SCE) and C70 (0.43 V vs. SCE). Solvent polarity effects On laser excitation of C70 in the presence of MOEP in toluene, the transient absorption spectrum exhibits the absorption band
of 3C70* at 980 nm without the appearance of MOEP + and C70 , suggesting that no electron transfer occurs. It was found that the initial concentration of 3C70* decreased with increasing MOEP concentration. The decays of 3C70* are linearly dependent on [MOEP], from which the kq values for the 3 C70*–MOEP systems were evaluated as (4.8–5.3) 109 mol1 dm3 s1 for M ¼ ZnII, MgII, (V=O)II and CuII. These kq values for 3C70*–MOEP systems in toluene are rather larger than those in BN. Triplet energy transfer from 3C70* to MOEP in toluene is also implausible, since the absorption band of 3MOEP* was not observed; this is reasonable because ET(3C70* ¼ 1.54 eV) is slightly lower than ET(3MOEP* ¼ 1.60–1.90 eV). Thus, the formation of the triplet exciplex 3[C70(d)–MOEP(d+)]* would be expected to be dominant in nonpolar solvent. The formation of such triplet exciplexes in nonpolar solvent has been reported during the quenching of the triplet states of metalloporphryrins by various quinones in toluene.6,7,36–39 Thus, the formation of 3[C70(d)–MOEP(d+)]* could provide a possible explanation for the near diffusion-controlled triplet quenching rate constant of 3C70* in the presence of MOEP in toluene as shown in Scheme 3. The k1 value may correspond to the observed kq . Kinetics analysis of the 3C70*–MgOEP system in mixtures of toluene–BN affords valuable information about the dissociation of 3[C70(d)—MOEP(d+)]* into the solvent-separated ionpair (SSIP) or into free ion radicals in solution. In the region of toluene-rich content (toluene > 75%), a transient triplet state absorption spectrum was observed, from which the kq values were obtained in a manner similar to that in toluene. Thus, it is presumed that less polar solvents retard the dissociation of the triplet exciplex into SSIP or into free ion radicals in solution.40 In the region BN > 25%, the dissociation of the triplet exciplex was confirmed by observing the absorption bands of C60 /C70 and MOEP +. This is reasonably interpreted by the stabilization of the SSIP and free ion radicals in a polar medium.40–42 In polar solvents, on assuming that all 3C70* pass through the triplet exciplex with MOEP, the Fet and ket values can be evaluated in a similar manner to those in BN, as listed in Table 3.
Scheme 2
Table 2 The kq , Fet and ket values for 3MOEP* with C60/C70 and the kbet values for C60 /C70 with MOEP + in Ar-saturated BN using 550 nm laser light System
kq/mol1 dm3 s1
Feta
ket/mol1 dm3 s1
kbet /mol1 dm3 s1
3
3.2 109 3.7 109 3.1 109 3.2 109 3.3 109 3.1 109 2.0 109 2.0 109
0.47 0.28 0.21 0.19 0.60 0.49 0.40 0.25
1.5 109 1.0 109 6.5 108 6.0 108 2.0 109 1.5 109 8.0 108 5.0 108
7.3 109 8.4 109 9.9 109 4.5 109 3.3 109 8.9 109 4.8 109 4.5 109
PdOEP*/C60 ZnOEP*/C60 3 MgOEP*/C60 3 (V=O)OEP*/C60 3 PdOEP*/C70 3 ZnOEP*/C70 3 MgOEP*/C70 3 (V=O)OEP*/C70 3
a
These values were calculated employing the reported molar extinction coefficients of 3MOEP* (47 000 mol1 dm3 cm1 at 440 nm).
3326
Phys. Chem. Chem. Phys., 2002, 4, 3322–3329
Table 3 The kq , Fet and ket values for 3C70*–(V=O)OEP and the kbet values C70 –(V=O)OEP + in Ar-saturated benzonitrile (BN)–toluene (tol) mixtures Solvents (BN : tol)
kq/mol1 dm3 s1
Fet
ket/mol1 dm3 s1
kbet/mol1 dm3 s1
25 : 75 50 : 50 60 : 40 75 : 25 87 : 13 100 : 0
3.3 109 3.3 109 3.3 109 3.0 109 2.8 109 2.6 109
— 0.25 0.30 0.35 0.39 0.40
— 8.3 108 9.8 108 1.1 109 1.1 109 1.1 109
— 9.7 109 7.4 109 5.6 109 5.3 109 4.5 109
Back electron transfer
Electron-mediating process to viologen dication (OV2+)
Fig. 9 shows the time profile of C70 on a long time-scale after the laser pulse in Ar-saturated BN solution. It is clearly shown that C70 begins to decay slowly after reaching the maximal absorbance. The decay time profile was fitted with secondorder kinetics, suggesting that bimolecular back electrontransfer from C60 /C70 to MgOEP + takes place. The bimolecular recombination process is considered to take place after solvation into the free radical ions, because first-order kinetics would be anticipated for the contact ion-pair. On employing the reported eA (C60 /C70 ), the kbet values were obtained as listed in Tables 1 and 2. The kbet values seem to be close to the diffusion controlled limit in BN. In a toluene–BN mixture, the decay was also fitted with second-order kinetics (Fig. 9), suggesting that the ion radicals are present as free ion radicals (eqn. (2)) or solvent-separated ion pairs (SSIP) (eqn. (3)). In the case of a contact ion pair (CIP) (eqn. (4)), the decay should be quick and obey first-order kinetics. The evaluated kbet values seem to increase slightly with increasing the toluene fraction, as listed in Table 3. A possible rationale for this finding involves the lower
viscosity of the toluene–BN mixture compared to BN. It can also be considered that the fraction of SSIP increases with toluene fraction, resulting in the increase in kbet values, since the kbet values in the SSIP are larger than those for free radical ions.
An electron-mediating process can be proved by the excitation of C70 (0.1 103 mol dm3) with MOEP (0.1 103 mol dm3) in the presence of an appropriate electron acceptor such as OV2+ (0.5 103 mol dm3), as shown in Fig. 10. The transient spectrum observed at 0.5 ms exhibits the characteristic bands of 3C70* at 980 nm, C70 at 1380 nm and MgOEP + at 650 nm. In the spectrum at 5 ms, the presence of OV + was observed by build up of absorption at 600 nm43 that parallels a concomitant decrease in the decay of C70 at 1380 nm. These observations indicate that, in the presence of MgOEP and OV2+, laser irradiation of C70 induces, at first, electron transfer from MgOEP to 3C70*. The C70 produced successively donates its excess electron to OV2+ yielding OV +, as indicated by the decay of C70 and rise of OV +. A possible rationale for this finding involves the lower reduction potential of OV2+ (0.3 V vs. SCE)44 compared to C70 (0.43 V vs. SCE).16–23 In this case C70 acts as an electron mediator in addition to being a photosensitizing electron acceptor. The photosensitized electron-transfer/electron-mediating processes are summarized in Scheme 1. It was found that the rate of OV + formation increases with increasing [OV2+], from which the electron mediating rate-constant was evaluated as 1.5 1010 mol1 dm3 s1 by applying the pseudo-first order plot. Furthermore, in long time-scale measurements, OV + at 600 nm begins to decay after reaching a maximum. The decay time-profile was fitted with secondorder kinetics, suggesting that final back electron-transfer (kfbet) takes place from OV + to MOEP +, regenerating OV2+ and MOEP. From the slope of the second-order plot for the long time decay profile of OV +, the final back electron transfer rate constant evaluated as 3.4 108 mol1 dm3 s1 using the reported extinction coefficient value of OV +.43 For MgOEP +, the decay of the radical cation in the presence of OV2+ is slowed down compared with those in the presence of C70 . This means that the lifetimes of the ion radicals of MgOEP +–OV + is quite longer than those of
Fig. 9 Decay of C70 at 1380 nm on a long timescale in the presence of MgOEP +; (a) in BN and (b) in BN : tol ¼ 50 : 50. Inset: Secondorder plots.
Fig. 10 Transient absorption spectra observed by 470 nm laser excitation of C70 (0.1 103 mol dm3) with MgOEP (0.1 103 mol dm3) in the presence of OV2+ (0.5 103 mol dm3) in Ar-saturated BN. Inset: Time profiles at 600 and 1380 nm.
In polar solvent;
ðC60 =C70 Þsolv þ ðMOEP þ Þsolv kbet 2nd-order ! C60 =C70 þ MOEP
In less-polar solvent;
In nonpolar solvent;
ðC60 =C70 ==MOEP ÞSSIP kbet 2nd-order ! C60 =C70 þ MOEP
ð2Þ
þ
ð3Þ
þ
ðC60 =C70 ; MOEP ÞCIP kbet 1st-order ! C60 =C70 þ MOEP
ð4Þ
Phys. Chem. Chem. Phys., 2002, 4, 3322–3329
3327
dissociation of the contact ion-pair (CIP) to solvent-separated ion pair (SSIP) or free ion radical is promoted. On addition of a viologen dication, the photosensitized electron-transfer/electron-mediating cycle was established, producing prolonged lifetimes of the holes of MOEP +. In the system MOEP–C60/C70–aromatic amine (DTQH), as a second donor having lower oxidation potential than MOEP, the hole shift from the radical cation of MOEP to the aromatic amine reagent was observed establishing a photosensitized electrontransfer/electron-mediating cycle.
Acknowledgement This work was partly supported by the Grant-in-Aid on Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan. Fig. 11 Transient absorption spectra observed by 550 nm laser excitation of MgOEP (0.1 103 mol dm3) with C70 (0.1 103 mol dm3) in the presence of DTQH (5 103 mol dm3) in Ar-saturated BN. Inset: Time profiles at 650 and 1020 nm. +
MgOEP –C70 . This may be caused by the electrostatic repulsion between the positive species that is expected to slow down back electron transfer.
References 1 2 3 4
Hole-shift cycle
5
In order to investigate the hole-shift process of MOEP +, one of the aromatic hydrazone derivatives (DTQH) was employed. Fig. 11 shows the transient absorption spectrum observed by 550 nm laser excitation of MgOEP (0.1 103 mol dm3) in the presence of C70 (0.1 103 mol dm3) and DTQH (5 103 mol dm3) in BN. The observed spectrum at 0.5 ms shows the characteristic bands of 3MgOEP* at 440 nm and MgOEP + at 650 nm. In the spectrum at 5 ms, the absorption band of C70 at 1380 nm was observed. Interestingly, the rise curve of DTQH + was observed at 510 and 1020 nm, which parallels a concomitant decay of MgOEP + at 650 nm, indicating that the hole-shift process takes place from MgOEP + to DTQH generating DTQH +. Scheme 2 summarizes the whole processes where the photosensitized electron transfer occurs at first from 3MgOEP* to C70 yielding C70 and MgOEP +, followed by the hole-shift from MgOEP + to DTQH yielding DTQH +. Such a hole shift is possible when the oxidation potential of the hole-shift reagent DTQH (Eox ¼ 0.32 V vs. SCE) is lower than that of MgOEP (Eox ¼ 0.53 V vs. SCE).28 Indeed for aromatic amines with higher Eox than 0.53 V, no charge shift was observed. The kfbet was evaluated as 6.1 108 mol1 dm3 s1 by following the long time decay profiles of DTQH + and C70 , which obey second-order kinetics. For C70 , the decay in the presence of DTQH + (kfbet ¼ 6.1 108 mol1 dm3 s1) is slowed down compared with those in the presence of MOEP + (kbet ¼ 4.7 109 mol1 dm3 s1). This means that the lifetime of the ion radicals of C70 –DTQH + is 10 times longer than that of C70 –MgOEP +.
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Conclusion A flash photolysis study shows that the ET process takes place via 3C70*/3C60* and 3MOEP* in Ar-saturated BN. The tendency of electron transfer rates and efficiencies via 3 C60*/3C70* with MOEP is seen to correlate nicely with the Gibbs energy changes (DGet0). A remarkable feature is that the 3C70* lifetime in the presence of MOEP was strongly reduced, even in toluene, without any evidence of energy transfer and/or electron transfer processes. On addition of polar solvents to nonpolar solvents, 3328
Phys. Chem. Chem. Phys., 2002, 4, 3322–3329
23 24 25 26 27 28
G. J. Kavarnos and N. J. Turro, Chem. Rev., 1986, 86, 401. Fundamentals of Photoinduced Electron Transfer, ed. G. J. Kavarnos, VCH, New York, 1993, ch. 3, pp. 103–184. D. M. Guldi, P. Neta and K.-D. Asmus, J. Phys. Chem., 1994, 98, 4617. C. S. Foote, in Physics and Chemistry of the Fullerenes, ed. K. Prassides, Kluwer, Dordrecht, 1994, pp. 79–96. Y.-P. Sun, in Organic Photochemistry, ed. V. Ramamurthy, and K. S. Schanez, Marcel Dekker, New York, 1997, pp. 325–390. J. Fujisawa, O. Yasunori and S. Yamauchi, Chem. Phys. Lett., 1998, 294, 248. M. Olmstead, D. A. Costa, K. Maitra, B. C. Noll, S. L. Phillips, P. M. van Calcar and A. L. Balch, J. Am. Chem. Soc., 1999, 121, 7090. D. M. Guldi and P. V. Kamat, in Fullerenes, Chemistry, Physics, and Technology ed. K. M. Kadish and R. S. Ruoff, Wiley-Interscience, New York, 2000, pp. 225–281. D. M. Martino and H. Willigen, J. Phys. Chem. A, 2000, 104, 10 701. S. Fukuzumi, in The Porphyrin Handbook, ed. K. Kadish, K. M. Smith and R. Guilard, Academic Press, San Diego, CA, 2000, vol. 8, pp. 115–151. K. M. Kadish, E. V. Caemelbecke and G. Royal, in The Porphyrin Handbook, ed. K. Kadish, K. M. Smith and R. Guilard, Academic Press, San Diego, CA, 2000, vol. 8, pp. 1–114. R. G. Kepler and P. A. Cahill, Appl. Phys. Lett., 1993, 63, 1552. Y. Wang, R. West and C. -H. Yaun, J. Am. Chem. Soc., 1993, 115, 3844. K. Takahashi, K. Etoh, Y. Tsuda, T. Yamaguchi, T. Komura, S. Ito and K. Murata, J. Electroanal. Chem., 1997, 426, 85. A. Ikeda, T. Hatano, S. Shinkai, T. Akiyama and S. Yamada, J. Am. Chem. Soc., 2001, 123, 4855. T. Nojiri, M. M. Alam, H. Konami, A. Watanabe and O. Ito, J. Phys. Chem. A, 1997, 101, 7943. M. E. El-Khouly, S. D-M. Islam, M. Fujitsuka and O. Ito, J. Porphyrins Phthalocyanines, 2000, 4, 713. M. E. El-Khouly, M. Fujitsuka and O. Ito, J. Porphyrins Phthalocyanines, 2000, 4, 590. T. Nojiri, A. Watanabe and O. Ito, J. Phys. Chem. A, 1998, 102, 5215. M. E. El-Khouly, Y. Araki, M. Fujitsuka and O. Ito, Photochem. Photobiol., 2001, 74, 22. C. Luo, M. Fujitsuka, T. Akasaka and O. Ito, J. Phys. Chem. A, 1998, 102, 8716. A. Masuhara, M. Fujitsuka and O. Ito, Bull. Chem. Soc. Jpn., 2000, 73, 2199. H. Onodera, Y. Araki, M. Fujitsuka, S. Onodera, O. Ito, F. Bai, M. Zheng and J-L. Yang, J. Phys. Chem. A., 2001, 105, 7341. D. Dubois, K. Kadish, S. Flanagan, R. Havfler, L. Chibante and L. Wilson, J. Am. Chem. Soc., 1991, 113, 4364. M. M. Alam, A. Watanabe and O. Ito, J. Photochem. Photobiol. A, 1997, 104, 59. S. I. Murov, Handbook of Photochemistry, Marcel Dekker, New York, 1995. G. Tollin, K. K. Chatterjee and G. Green, Photochem. Photobiol., 1995, 4, 593. K. M. Kadish, K. M. Smith, R. Cuilard, The Porphyrin Handbook, Academic Press, New York, 2000, vol. 9, p. 117.
29 V. S. Chirvonyi, B. M. Dzhagarov, Y. V. Timinskii and G. P. Gurinovich, Chem. Phys. Lett., 1980, 70, 79. 30 K. Bohandy and B. F. Kim, J. Chem. Phys., 1983, 78, 4331. 31 E. F. Hilinski, K. D. Straub and P. M. Rentzepis, Chem. Phys. Lett., 1984, 111, 333. 32 J. Darwent, P. Douglas, A. Harriaman, G. Porter and M. C. Richoux, Coord. Chem. Rev., 1982, 44, 83. 33 L. Biczok, H. Linschitz and A. Treinin, in Fullerenes, ed. K. M. Kadish and R. S. Ruoff, the Electrochemical Society, Pennington, NJ, 1994, vol. 94-24, p. 909–920. 34 T. Konishi, M. Fujitsuka, O. Ito, Y. Toba and Y. Usui, J. Phys. Chem. A., 1999, 103, 9938. 35 D. M. Guldi and M. Prato, Acc. Chem. Res., 2000, 33, 695. 36 J. K. Roy, F. A. Cattol and D. G. Whitten, J. Am. Chem. Soc., 1972, 94, 7164.
37 J. K. Roy, F. A. Cattol and D. G. Whitten, J. Am. Chem. Soc., 1974, 96, 6349. 38 A. Harriman, G. Porter and N. Searle, J. Chem. Soc. Faraday Trans. 2, 1975, 75, 1515. 39 W. Boyd, P. D. Hodgson, M. C. Rickard, C. E. F. Oliver, A. G. Chaker, L. Brothers, P. J. Bolskar, F. S. Tham and R. D. Reed, J. Am. Chem. Soc., 1999, 121, 10 487. 40 N. Mataga, T. Okada and N. Yamamoto, Chem. Phys. Lett., 1967, 1, 119. 41 Y. Taniguchi and N. Mataga, Chem. Phys. Lett., 1972, 13, 596. 42 H. Hase and Y. Miyataka, Chem. Phys. Lett., 1993, 215, 141. 43 S. Ferrere, A. Zaban and B. A. Gregg, J. Phys. Chem. B, 1997, 101, 4490. 44 J. Peon, X. Tan, J. D. Hoerner, C. Xia, Y. F. Luk and B. Kohler, J. Phys. Chem. A, 2001, 105, 5768.
Phys. Chem. Chem. Phys., 2002, 4, 3322–3329
3329
