Electron Transfer in Oligothiophene-Bridged ... - ACS Publications

Feb 3, 2009 - für Chemie, Max-Volmer Labor, Technische UniVersität Berlin, Strasse des 17. .... 1.0 at the maximum of the absorption Q-band with the...
1 downloads 0 Views 711KB Size
Download PDF
2526

J. Phys. Chem. B 2009, 113, 2526–2534

Electron Transfer in Oligothiophene-Bridged Bisporphyrins Martin Regehly,† Tianyu Wang,§,⊥ Ulrich Siggel,*,‡ Ju¨rgen H. Fuhrhop,§ and Beate Ro¨der*,† Institut fu¨r Physik, Photobiophysik, Humboldt UniVersita¨t, Newtonstr. 15, D-12489 Berlin, Germany; Institut fu¨r Chemie, Max-Volmer Labor, Technische UniVersita¨t Berlin, Strasse des 17. Juni, D-10623 Berlin; and Institut fu¨r Chemie und Biochemie, Organische Chemie, Freie UniVersita¨t Berlin, Takustr. 3, D-14195 Berlin, Germany ReceiVed: September 10, 2008; ReVised Manuscript ReceiVed: December 2, 2008

Oligothiophene-bridged zinc-tin bisporphyrinates were synthesized. Their absorption spectra have been analyzed in terms of exciton interaction and porphyrin-bridge coupling by through-bond interaction and the steady-state fluorescence spectra in terms of differential Stokes shifts for the electron-donating zinc and the electron-accepting tin porphyrinates. Strong quenching of the fluorescence intensity and acceleration of the fluorescence decay as compared to porphyrinate monomers (ZnTPP, SnTPP) were observed. Both phenomena were traced back to light-induced electron transfer by the occurrence of ion pair absorption bands in picosecond transient absorption spectra. Similar absorption spectra of both chromophores caused always simultaneous excitation and, consequently, two concurrent photoreactions. Combined evaluation of the time-dependent absorption and fluorescence data allowed the estimation of rates for the electron transfer reactions. The found dependence on the separation distance was much smaller than for donor-acceptor systems with saturated spacers. A damping factor of 0.05 was calculated for the charge separation proceeding from the excited state of the zinc porphyrin. The polarity of the solvent had a profound influence on the transfer rates. The charge recombination was 300 times faster in polar tetrahydrofuran than in nonpolar toluene. 1. Introduction Photoactive bridged donor-acceptor systems (DAB systems) have been synthesized for different reasons in the past years. One of the initial motives has been to mimic photosynthesis in realizing long-lived ion pair (IP) states.1,2 After efforts with flexible bridges3 the introduction of rigid bridges made it feasible to check the free energy dependence of electron transfer rates with fixed bridge within the framework of the Marcus theory,4 on the one hand, and the dependence on the length of the bridge with constant driving force, on the other hand.5,6 Later it was found that by substituting aliphatic for aromatic spacers the distance dependence, which up to then made electron transfer slow or even impossible at larger values of the distance, could be significantly decreased.7 The bridge is not only a spacer but mediates the coupling between donor and acceptor by throughbond interactions (superexchange mechanism).8 Helms et al.9 found a value of 0.4 Å-1 for the damping factor β with polyphenylene bridges; Petterson et al.10 found β ) 0.29 Å-1 with phenyleneethynylene bridges. In the search of highly effective bridges for maintaining high rates of electron or energy transfer despite appreciable distances between donor and acceptor, oligothiophenes have been introduced,11 generally known as molecular wires.12 Thiophene with a weaker resonance energy relative to benzene favors long-range electron delocalization. The group Effenberger-Wolf13a,b has used oligothiophenes as mediators of effective energy transfer * Corresponding authors: Tel +49 30 2093 7625; Fax +49 30 2093 7666; e-mail [email protected] (U.S.), [email protected] (B.R.). † Humboldt Universita¨t. ‡ Technische Universita¨t Berlin. § Freie Universita¨t Berlin. ⊥ Present address: CAS Key Laboratory of Colloid, Interface and Chemical Thermodynamics, Institute of Chemistry, The Chinese Academy of Sciences, Beijing 100080, P.R. China.

SCHEME 1: Chemical Structures of T-1, T-2, and T-3 and Reference Compound ZnT-3

between an anthryl group and a free base porphyrin. And energy transfer likewise has been realized between a zinc and a free base porphyrin by the Quintard-Hammarstro¨m group,14 using simple and modified oligothiophenes as bridges. We aimed at realizing electron transfer and selected a zinc porphyrin as donor and a tin porphyrin as acceptor for thermodynamic reasons. Both metal complexes were connected by oligothiophenes of different length (see Scheme 1). In particular, it had been shown earlier that photoexcited tin(IV) porphyrinates are reduced to phlorins and evolve molecular hydrogen in presence of weak reductants.15 Tin porphyrins could thus play the role of the acceptor in photosynthesis, the zinc porphyrinate that of a chlorophylla analogue, and the thiophenes play the membrane protein role of electron conductors. In spite of the difficulties arising from similar absorbance and fluorescence spectra of the two porphyrins, we have been able to prove electron transfer to occur after photoexcitation. We could show that despite nonplanarity of the system because of steric hindrance there is effective electronic coupling between the

10.1021/jp808052u CCC: $40.75  2009 American Chemical Society Published on Web 02/03/2009

Oligothiophene-Bridged Bisporphyrins

J. Phys. Chem. B, Vol. 113, No. 8, 2009 2527

SCHEME 2: Synthesis of T-1, T-2, and T-3

porphyrins, at least in a polar medium, which leads to a weak dependence of the electron transfer rate on the distance. 2. Experimental Section 2.1. Analytical Methods and Preparation of the Samples. 1 H NMR spectra were recorded on a Bruker AC 270 (operating at 270 MHz), chemical shifts being reported in ppm relative to Me4Si (TMS). 13C NMR spectra were recorded on an AC 500 instrument (at 500 MHz). Mass spectra were recorded on a Finnigan MAT 711. The positive FAB (fast atom bombardment) ionization method was used. Compounds T-1, T-2, and T-3 were synthesized in a stepwise procedure according to Scheme 2. The details of synthesis can be found in the Supporting Information. 2.2. Absorption and Steady-State Fluorescence. The groundstate absorption spectra were recorded using a commercial spectrophotometer (Shimadzu UV-2501PC) at room temperature. Steady-state fluorescence spectra of the investigated compounds were measured in 1 cm × 1 cm quartz optical cells using a combination of a CW xenon lamp (XBO 150) and a monochromator (Lot-Oriel, bandwidth 10 nm) for excitation and a polychromator with a cooled CCD matrix as a detector system (Lot-Oriel, Instaspec IV).17 2.3. Time-Resolved Fluorescence. Fluorescence lifetimes were measured by the time-correlated single photon counting

(TCSPC) technique, using the frequency-doubled pulses of a Ti:sapphire laser (Coherent Mira 900, 405 nm, fwhm 200 fs) for excitation. The instrument response function was 33 ps, as measured at excitation wavelength with Ludox. A description of the setup has been previously published.18 A self-made routine was applied to minimize the least-squares error between the model function convoluted with instrument response function and the measured data set. 2.4. Picosecond Transient Absorption Spectroscopy (psTAS). To measure transient absorption spectra, a white light continuum was generated as a test beam in a cell with D2O/ H2O mixture using intense 25 ps single pulses from a Nd:YAG laser (PL 2143A, Ekspla) at 1064 nm. Before passing through the sample, the continuum radiation was split to get a reference spectrum. The transmitted as well as the reference beams were focused into two optical fibers and were recorded simultaneously at different traces on a CCD matrix (Lot-Oriel, Instaspec IV). A tunable radiation from OPO/OPG (Ekspla PG 401/SH, tuning range 200-2300 nm) pumped by the third harmonic of the same laser was used as an excitation beam. The mechanical delay line allowed the measurement of the light-induced changes in the absorption spectrum at different time delays up to 15 ns after excitation.19 The optical density (OD) of all samples was 1.0 at the maximum of the absorption Q-band with the longest wavelength.

2528 J. Phys. Chem. B, Vol. 113, No. 8, 2009

Figure 1. Absorption spectra of T-1, T-2, and T-3 and reference compound ZnT-3 in THF.

2.5. Molecular Modeling. In an initial step the structures of T-1, T-2, and T-3 (see Scheme 1) were geometry optimized using the module Discover of MS Modeling. For this purpose a COMPASS (condensed-phase optimized molecular potentials for atomistic simulation studies) force field was chosen. In the next step geometry optimization was refined by density functional theory employing the DMOL3 package of MS Modeling. Calculation was performed with the local density approximation (LDA) and the double numerical basis set including dpolarization (DND). All electrons are included in the calculation. The convergence thresholds for energy change, maximum force, and maximum displacement had been set to 2 × 10-5 hartree, 0.004 hartree/Å, and 0.005 Å, respectively. Geometry optimization stopped whenever the energy convergence criteria are satisfied along with either the gradient or displacement criteria.20 3. Results 3.1. Synthesis. The thiophene-bridged bisporphyrins T-1, T-2, and T-3 were prepared by a Suzuki cross-coupling of a meso-boronated diarylporphyrin 2a and R-brominated thiophenes,21 which were bound to a methine bridge of the second porphyrin 3a,b′,c′. These triaryloligothiophenylporphyrins were obtained by a variation of the Rothemund method.22-24 Four parts of the unsubstituted pyrrole were condensed oxidatively with 3 parts of mesitylaldehyde and 1 part of the carbaldeyde of mono-, bis-, or terthiophene using boron trifluoride etherate as a catalyst. The diarylporphyrin 2 was synthesized in two steps. First, the substituted dipyrromethane 1 was prepared by condensation of pyrrole with 4-(2-ethylhexyloxy)benzaldehyde.25,26 In the second step a variation of the MacDonald method26,27 was adopted. Two molecules of dipyrromethane were condensed oxidatively with trimethyl orthoformate (yielding the methane bridges) and trichloracetic acid to give the 5,15-bis(4-(2-ethylhexyloxy)phenyl)porphyrin 2. Three more modifications of the porphyrin, namely bromination, metalation with zinc acetate, and reaction with pinacolborane and palladium bis(triphenylphosphin) dichloride as a catalyst, yielded the porphyrin boronate 2a, which was finally coupled to the bromooligothiophene porphyrins in the presence of palladium-tetra(triphenylphosphine) as a catalyst to yield the bridged bisporphyrins 4a, 4b, and 4c. The free base moieties were finally metalated either with tin(II) chloride or with zinc acetate to give T-1, T-2, and T-3 and the reference compound ZnT-3. We have chosen two porphyrins with different substitution for the bisporphyrin for reasons of solubility. The two isooctyl chains prevent aggregation. 3.2. Absorption. The absorption spectra of T-1, T-2, and T-3 and compound ZnT-3 in THF are shown in Figure 1. ZnT-3 is used as reference compound, as the porphyrin ligands are identical to those in T-1, T-2, and T-3. The maxima of the

Regehly et al.

Figure 2. Normalized fluorescence spectra of T-1, T-2, and T-3 and reference compound ZnT-3 in THF.

TABLE 1: Absorption Data of Synthesized Bisporphyrins and Comparable Monomers compound T-1 T-2 T-3 ZnT-3 ZnTPP SnTPP Σ(ZnTPP+ SnTPP)

Soret fwhm Soret Q(1,0) Q(0,0) fwhm Q(0,0) band [nm] band [nm] [nm] [nm] band [nm] 423/437 426/430 426 427 423 429 424/428

26 25 22.7 22.5 9.3 6.6 14

563 559 559 560 555 563 557/561

609 607 607 605 595 602 601

32 40 (40) (34) 20 9 18

TABLE 2: Position of Maxima (p) and Shoulders (sh) in the Fluorescence Spectra and Relative Fluorescence Quantum Yields compound

λshort emission [nm]

λlong emission [nm]

far red emission [nm]

T-1 T-2 T-3 ZnT-3

630 (sh) 634 (p) 633 (sh) 635 (p)

661 (p) 656 (sh) 653 (p) 662 (sh)

725 718

ΦFrel 0.05 0.14 0.08 1

absorption of these compounds are listed in Table 1 together with those of ZnTPP and SnTPP. The Soret bands of the bisporphyrins are much broader (half-width 23-26 nm) than those of the monoporphyrins (half-width 6.6 nm for SnTPP and 9.3 nm for ZnTPP) and also than that of the 1:1 mixture of them (half-width 14 nm). T-1 with the shortest distance between the porphyrin units has a split Soret band, whereas in T-2 the splitting is hardly seen and in T-3 and ZnT-3 not at all. The Q-bands of all the four bisporphyrins are very similar. The Q(1,0) bands around 560 nm are approximately the sum of ZnTPP and SnTPP absorptions with some 30% broadening. The Q(0,0) bands around 607 nm are subject to a much greater change. They are clearly red-shifted and characterized by a long tail up to 700 nm and a half-width 2 times as large as that of the 1:1 mixture of ZnTPP and SnTPP (Table 1 and Supporting Information). 3.3. Steady-State Fluorescence. The fluorescence of T-1, T-2, and T-3 is strongly quenched compared to ZnT-3. The relative quantum yields given as the ratio of areas under the pertinent fluorescence spectral curves are shown in Table 2. The quenching is highest (5% yield only) for T-1 with the shortest bridge length, but the lowest quenching (14% yield) is realized for T-2 with an intermediate bridge. The fluorescence spectrum of reference ZnT-3 (Figure 2) is different from what is observed for regular monomeric porphyrins as MgOEP or ZnTPP. It consists of a broad band with about 75 nm half-width. This value is also found for the overall band of regular porphyrins, if the gap between the usually two peaks is disregarded. The peak of the spectrum is located at 635 nm;

Oligothiophene-Bridged Bisporphyrins

J. Phys. Chem. B, Vol. 113, No. 8, 2009 2529

Figure 3. Fluorescence decay of T-1, T-2, and T-3 in THF and T-2 in toluene. The instrument response function (IRF) is shown as a dashed gray line.

TABLE 3: Fluorescence Decay Times τi and Their Respective Amplitudes ai for Excitation at 400 nm and Detection at 630 nm compound

τ1 [ps] (a1 [%])

τ2 [ps] (a2 [%])

τ3 [ns] (a3 [%])

τ4 [ns] (a4 [%])

T-1 T-2 T-3 T-2 toluene

17 (73) 28 (69) 26 (74) 220 (31)

260 (11) 230 (13) 160 (18) 960 (68)

1.2 (12) 1.1 (16) 1.1 (6)

4.8 (3) 4.6 (2) 3.8 (2) 3.4 (1)

i.e., the Stokes shift of 30 nm is much larger than for ZnTPP (12 nm). The spectrum is also slightly dependent on the excitation wavelength, indicating that the two ZnP chromophores do not have the same electronic properties. Excitation at 475 nm (where there is a shoulder on the long wavelength slope of the Soret band) leads to a fluorescence band with a maximum at 648 nm. The dependence on the excitation wavelength is stronger for compounds T-1, T-2, and T-3 (not shown here). The fluorescence of T-1, T-2, and T-3 differs not only in their respective intensity but also in the spectrum (Figure 2). T-2 and the reference compound ZnT-3 have nearly identical emission spectra with a maximum at 635 nm when the excitation wavelength was set at 405 nm. T-1 and T-3 exhibit a shoulder near this wavelength, whereas the maximum is located at 660 and 653 nm, respectively. 3.4. Time-Resolved Fluorescence. The fluorescence decay of compounds T-1, T-2, and T-3 is not monoexponential and has been fitted by four exponential functions (Figure 3a-d). The dominant fraction of the decay (between 80% and 90%) is characterized by two times having average values of 24 and 217 ps. These decay components are relatively similar for T-1, T-2, and T-3; the individual values together with the corresponding relative amplitudes are given in Table 3. A minor portion of the fluorescence decays much more slowly with lifetimes of 1.1 ns (about 12% relative amplitude) and 4.4 ns (2%). These two fractions are attributed to impurities which are probably due to the limited stability of the bisporphyrins. It should be noted that the time-resolved fluorescence does not show a clear dependence on the length of the bridge, which consists of 1, 2, and 3 thiophene units for the three molecules, respectively. The very fast decay of fluorescence compared to that of the monomers ZnTPP and SnTPP described above has been measured with THF as a polar solvent. In a nonpolar solvent the decay is much slower. For T-2 in toluene 99% of the decay can be described by two exponentials with lifetimes of 220 and 960 ps, which are 8 and 4 times longer than in THF. 3.5. Transient Absorption Spectroscopy. The facts that the steady-state fluorescence is strongly quenched and the fluores-

Figure 4. Selected transient absorption spectra of T-2 and T-3 at different delay times.

cence decay after laser excitation is accelerated were further followed by transient absorption spectroscopy in order to look for products of a possible light-induced charge separation. Compound T-1 could not be investigated because it was photochemically too unstable. Compound T-2 has been examined in THF and toluene, since there are hints in the literature that products may be more easily identified in nonpolar solvents.28 T-2 was examined in both solvents, THF and toluene, and excited at 562 nm, and T-3 only in THF at 550 nm. The measured transient absorption spectra are shown in Figure 4 for different time delays. The spectra after a delay time of 50 ps for THF and 3 ns for toluene as a solvent are similar, but not identical, for the three samples. They are characterized, apart from positive absorption changes, by a decrease of absorption around 605 and 560 nm, which are the wavelengths of the Q-bands, the Q(0,0) transition, and its vibronic overtone Q(1,0). The positive absorption changes between 640 and 900 nm, where the ground-state absorption is low to negligible, are ascribed to a radical pair, formed by charge separation. There is a maximum at 660 nm and a second one at 720-730 nm, which is most clearly seen in the transient spectrum measured for T-2 in toluene (Figure 4b). The radical cation of ZnTTP is known to have an unstructured broad absorption between 620 and 700 nm.29,30 The radical anions of tin porphyrins such as SnTPyP and SnTPPS4- absorb between 600 and 800 nm31 with a peak around 720 nm (εmax ) 1.2 × 104 M-1 cm-1). There are also positive absorption changes between 450 and 540 nm. The peak wavelength is shifted from 490 to 500 nm during the first 500 ps and from 500 to 520 nm for T-2 and T-3, respectively. In this region absorption due to S1-Sn and T1-Tn transitions is expected.28,32,33 For the compounds studied in the aforementioned references, the maximum of the transient absorption changes is, however, located at an appreciably smaller wavelength (460 nm), most probably because the zinc porphyrin macrocycle is only ineffectively coupled to an electron acceptor through a phenyl group. One result that needs to be clarified is the absence of absolutely negative absorption changes in the region were ground-state absorption is observed. The measured changes are only negative relative to their neighborhood. This is caused by the low absorption strength of the Q-bands and by the fact that they are the only sharp bands in the region. All the other

2530 J. Phys. Chem. B, Vol. 113, No. 8, 2009

Figure 5. Ground-state depletion and time evolution of radical formation as determined by TAS. Compilation of time constants and amplitudes (in brackets) from the curve fitting: (A) 3 exponentials with τ1 ) 28 ps (0.02), τ3 ) 150 ps (0.07), τ2 ) 320 ps (0.03) and constant c ) 0.02; (B) OD(t) ) A(exp(-t/τ3) - exp(-t/τ1)), with τ1 ) 43 ps, τ3 ) 150 ps, A ) 0.4; (C) 3 exponentials with τ1 ) 220 ps (0.02), τ2 ) 960 ps (0.05), τ3 ) 36 ns (0.14); (D) ∆OD(t) ) A(exp(-t/τ3) exp(-t/τ1)) + B(exp(-t/τ3) - exp(-t/τ2)), with τ1 ) 220 ps, τ2 ) 960 ps, τ3 ) 53 ns, A ) 0.06; B ) 0.14; (E) 2 exponentials with τ2 ) 160 ps (0.02), τ3 ) 66 ps (0.07), and constant c ) 0.01; (F) ∆OD(t) ) A(exp(-t/τ3) - exp(-t/τ1)), with τ1 ) 42 ps, τ3 ) 66 ps, A ) 1.03. The standard deviation of each measured data point was estimated to 10% and 20% of its value for S0 depletion and ∆OD(720 nm), respectively.

absorption bands due to excited states and electron transfer products are very broad, covering also the region of groundstate absorption. This has been observed earlier e.g. for a dyad of zinc porphyrin and pyromellithimide by the Mataga group28 and for a dyad of tetraphenylporphyrin and benzoquinone by the Netzel group.46 The time evolution of the absorption changes is exemplified for two wavelengths in Figure 5. At 720 nm (Figure 5B,F) the absorption increases relatively slowly to a maximum after 75 ps (T-2) or 50 ps (T-3), in order to decay to zero in the following 500 ps, in accordance with the notion that reaction products are monitored. The time course can approximately be fitted by two exponential functions. The characteristic time constants have been found to be 43 and 42 ps for the buildup and 150 and 66 ps for the decay of the products for T-2 and T-3, respectively. In toluene the processes are much slower (Figure 5D); the maximum of the absorption changes was attained at 3 ns. The result of the fitting procedure is a 2-fold increase of ∆OD with time constants of 220 and 960 ps and a common decay time of 53 ns. For λ ) 562 nm we have calculated the relative depletion of the ground state, which is equal to the relative number of molecules in the excited states, using a compensation method.34-36

Regehly et al. A portion of the ground-state absorption spectrum is added to the measured difference spectrum up to the point, where the negative changes, corresponding to the bleaching of the ground state, are eliminated. This portion is equal to the relative depletion under the condition that the excited state S1 (or T1 at a later time) and the products do not exhibit prominent transient absorption peaks in the range of Q-absorption of the ground state. This has been shown to be true explicitly for H2TTP and ZnTTP in ref 34 and is true for the radicals in question.29,31 Additionally, negative absorption changes due to induced emission have to be compensated. In our case they are negligible because of the low quantum yield of fluorescence. No negative absorption changes at the wavelength of maximum fluorescence around 625 nm have been detected. The results of the calculation are shown in Figure 5A,C,E. The ground-state depletion reaches its maximum at 30-50 ps, which corresponds to the half-width of the exciting laser flash of 25 ps. The decay has been fitted to the sum of up to three exponential functions and is different for the three samples. The corresponding time constants are 28, 150, and 320 ps for T-2, 66 and 160 ps for T-3, and 220 ps, 960 ps, and 36 ns for T-2 in toluene. Additionally, a portion of the depletion is irreversible within the maximal time interval of our measurements (15 ns). As the excited singlet state and the radical states are no more populated at this time, this portion has to be ascribed to the triplet state. 3.6. Molecular Modeling. The geometry optimization of T-1, T-2, T-3 and reference compound ZnT-3 was carried out in vacuum. The calculated structures with the lowest potential energy are shown in Figure 7. ZnT-3 is not depicted because the geometry of the molecule is very similar to T-3. No indication was found for multiple minima; i.e., different conformers do not seem to play a role. This is consistent with the results of Odobel et al.,14 where different conformers are not mentioned for bridges of two and four thiophene molecules. For T-1, T-2, and T-3 the center-to-center distances between zinc and tin porphyrin were determined to 12, 16, and 21 Å in that order. The angles between the Zn-porphyrin plane and the first thiophene unit of the bridge were estimated to 49°, 40°, and 43° for T-1, T-2, and T-3, respectively. The corresponding angles with the plane of the Sn-porphyrin are very similar (52°, 41°, and 46°).The distances and angles are listed in Table 4. Thus, the first thiophene ring which is responsible for the coupling between the porphyrin and the bridge is neither parallel nor perpendicular to the plane of the porphyrins; it holds an intermediate position of ∼45°, which should lead to intermediate coupling only. Additionally, the tilt angles between the ZnP and SnP transition moments (φ) were determined to 18°, 26°, and 21° for T-1, T-2, and T-3, respectively. The trimethylphenyl substitutents of the tin chromophore are aligned perpendicular to the plane of the porphyrin macrocycle. In comparison to this arrangement, the 2-ethylhexyloxyphenyl substitutents of the zinc porphyrin are turned out of the porphyrin plane to less extent. 4. Discussion 4.1. Ground-State Absorption and Steady-State Fluorescence. For the discussion of the properties of the compounds T-1 to T-3 it is essential to note that efficient conjugation through the system of double bonds is not expected. The thiophene rings are sterically hindered to adopt a position in which all the planes of the different rings are parallel. We will see that the bisporphyrins are no new molecules with totally new properties. The properties of the two porphyrins are partially retained and only modulated by interactions.

Oligothiophene-Bridged Bisporphyrins

J. Phys. Chem. B, Vol. 113, No. 8, 2009 2531

Figure 6. Simulation of fluorescence spectra of T-3 and reference compound ZnT-3 in THF. The spectrum of T-3 is split into a ZnP and a SnP spectrum with a peak wavelength of 630 and 653 nm, which is clearly red-shifted as compared to ZnTPP (602 nm) and SnTPP (603 nm). The spectrum of ZnT-3 is split into two ZnP spectra with peak wavelengths of 625 and 645 nm.

Figure 7. Geometry-optimized structures of T-1, T-2, and T-3 as calculated with density functional theory using the module DMOL3 of MS Modeling.

TABLE 4: Calculated Free Energy Changes ∆G° for Light-Induced Charge Separation for Compounds T-1, T-2, and T-3 in THF and Toluene, Using the Measured Redox Potentials of ZnTPP and SnCl2TPP, and Intramolecular Distances Ra compound T-1 T-2 T-3

∆G° ∆G° R R γ φ (THF) [eV] (toluene) [eV] [Å] [deg] [deg] [deg] -0.55 -0.51 -0.48

0.01 0.13 0.22

12 16 21

49 40 43

52 41 46

18 26 21

a Furthermore, the angles between the plane of the first thiophene unit and the planes of ZnP (R) and SnP (γ) are given, respectively. The tilt angles between ZnP and SnP transition moments (φ) are shown additionally.

The Soret absorption of the four dyads is characterized by the exciton interactions of the nearly degenerate Soret transitions of the ZnP and SnP chromophores. The interaction is strongest for compound T-1 with one thiophene unit as a bridge (12 Å center-to-center distance), leading to two Soret bands 14 nm apart. In T-2 the two peaks are only about 4 nm apart and not clearly separated. In the T-3 compounds with 21 Å distance broadening is the only indication for interaction (22.6 nm halfwidth). The broadening does not vary very much with the distance of the chromophores (see Table 1), on the one hand, but is astonishingly large as compared to the envelope around the Soret bands of ZnTPP + SnTPP (see Supporting Information). This points to a broadening by electronic through-bond interaction with the bridge. The Q-bands have much smaller transition moments. Therefore, appreciable exciton interaction is neither expected nor observed for all the compounds. The position of the Q-bands for the zinc and tin porphyrins which are 7 nm apart in the monomers is still discernible in the 1:1 mixture but not anymore visible in the bisporphyrins. As

compared to the envelopes around the corresponding bands of the mixture ZnTTP/SnTTP, the Q(0,0) bands are especially broadened (more than 30 nm half-width as compared to 17.5 nm; see Table 1 and Supporting Information), which is probably caused by interaction with the bridge as in the case of Soret broadening. Additionally, they are red-shifted by 5 nm (T-2, T-3) and 8 nm (T-1). This is interpreted as a solvent effect. Q(0,0) is known to be especially sensitive to the environment of the chromophore. In our case there are fewer solvent molecules near the chromophores in the bridged dyads, especially in T-1. The inspection of the measured fluorescence spectra has shown two peculiarities. The spectra are different from those of ZnTPP and SnTPP regarding their shape and their position. The notion that they have to be interpreted as charge transfer spectra had to be rejected for two reasons. There is no clearly defined charge transfer band in the absorption. If two identical porphyrins are bridged by thiophenes,14 they exhibit a much more normal, only red-shifted fluorescence spectrum, although charge transfer to the bridge occurs in the excited state. The second peculiarity is that T-2 shows a spectrum different from that of T-1 and T-3, but similar to that of ZnT-3, which is contrast to the very small differences in the Q(0,0) absorption bands of all the four compounds. Our interpretation of the fluorescence spectra is based on the experimental fact that the relative amplitudes of shoulders and peaks depend on the wavelength of excitation, which is true not only for the Zn-Sn bisporphyrins but also for the Zn-Zn compound. This suggests that the measured spectra represent the envelope above two emissions with some difference in the position, originating from the two porphyrin moieties. We then tried to compose the measured spectra from two monomer fluorescence spectra. For the zinc porphyrin part we took a spectrum published by Odobel et al.14 for a bisporphyrin of two identical Zn porphyrins coupled by a tetrathiophene bridge. It is characterized by a peak at 626 nm (i.e., there is a red shift of about 25 nm as compared to ZnTPP) and a high vibrational band, scarcely separated from the (0,0) transition band. This is the result of interaction with the bridge. We started with ZnT-3 and tried to compose its fluorescence spectrum of two Odobel spectra with some difference in peak wavelength. The result is shown in Figure 6B. The porphyrin with trimethyl substitution (ZnP2) has its maximum of fluorescence at 645 nm, i.e., by 20 nm at longer wavelength than ZnP1 with ethylhexyloxyphenyl substitution. Thus, we arrive at the astonishing result that two relatively similar zinc porphyrins show appreciably different Stokes shifts as result of interaction with the terthiophene bridge. The two spectra with an intensity ratio of 1:0.75 will be used to calculate the intensity quenching for the oxidative and reductive quenching reactions (see next section). For the synthesis of the fluorescence spectra of T-1, T-2, and T-3 we started similarly as for ZnT-3. We took the shape of

2532 J. Phys. Chem. B, Vol. 113, No. 8, 2009

Regehly et al.

the spectrum of Odobel et al.14 and fitted its position to the short wavelength edge of our measured spectra. We had still to make a choice for its amplitude. If we assume that the shortest fluorescence decay times are produced by the ZnP moiety and the second shortest by the SnP moiety, which will be proven later (see section 4.2), we can calculate the relative emission intensities for the ZnP and SnP moieties from the ratio of the corresponding amplitudes of time-resolved fluorescence. These amplitudes are proportional to the product of absorption coefficient at the wavelength of excitation and the rate constant of fluorescence. The intensity of steady-state fluorescence is additionally proportional to the lifetime of fluorescence. From the ratio of measured amplitudes a1/a2 of the two main decay processes of 6.64, 5.3, and 4.1 for T-1, T-2, and T-3, respectively, we calculated an expected ratio of integrated intensities I(Zn)/I(Sn) for the zinc and tin porphyrin moieties of 1:2.3, 1:1.55, and 1:1.5. The shape of the fluorescence spectrum of the tin porphyrin bound to an oligothiophene bridge is not known. We have therefore taken the difference of our measured spectra and the ZnP spectrum and have only chosen the area of this difference spectrum according to our calculation. For the spectra of T-1 and T-3 with their shoulder on the short wavelength side the simulation is successful with intensity ratios of 1:2.6 and 1:1.5, which is shown in Figure 6A for T-3. And for T-2 despite the different shape of the total emission spectrum, a composition has been found to be possible with a similar ratio for the areas of the individual emissions (1:1.54) as for T-3. The difference is now present in the SnP spectrum, in the relatively small Stokes shift (36 nm) and the absence of the second band around 700 nm. Looking at the positions of the spectra in our fit, we conclude that a Stokes shift of 18-30 nm for the ZnP moiety exists which is somewhat larger than that observed by Odobel et al.14 (25 nm). For the tin porphyrin moiety the Stokes shift is even larger (36-50 nm). It has to be stressed that we can give only approximative values for the Stokes shift. In the Q(0,0) absorption band the position of the SnP and ZnP part are not visible. Likewise, we have not examined if in contrast to the usual nomenclature there are vibronic bands under the Q(0,0) envelope. As has been discussed in a recent publication,47 this would be necessary if measured values are to be compared with those predicted by theories. 4.2. Electron Transfer. We have shown that in the T-n systems the fluorescence intensity is highly quenched and the fluorescence decay strongly accelerated in comparison to reference compounds. For an interpretation we check at first if the thermodynamic prerequisites for a light-induced charge separation are met. Two reactions have to be considered. Oxidative quenching of the excited state of the zinc porphyrin by the tin porphyrin in the ground state and reductive quenching of the excited state of the tin porphyrin by the zinc porphyrin in the ground state. The driving force can be calculated from ox red and E1/2 and the the pertaining measured redox potentials E1/2 energy ∆E* of the first excited state (0-0 transition) of the species involved. red + ∆G° ) e[Eox 1/2(D/D ) - E1/2(A/A )] - ∆E* -

[

][

]

1 e2 1 1 e2 1 + (1) 4πε0εsR 8πε0 rd+ ra- εref εs This equation is approximative in that sense that it treats the charge-separated state of the bisporphyrin as two spherical ions with radii ra- and rd+, separated by the distance R. Two correction terms are included in eq 1: The first is a solvation

term for the case that the measurement of the redox potentials has been carried out in a different solvent (with dielectric constant εref) than the photoreaction under study (solvent with εs). The second correction is a Coulomb energy term which takes into account that after charge separation the two resultant radicals of the bisporphyrins are at the fixed distance R and not at infinite distance as required for the redox potentials. As electrochemical data are not available for our compounds T-n we can only present an approximative calculation using halfwave potentials of ZnTTP and SnTPP. 37 Eox 1/2 (0.71 V) is the oxidation potential of the donor ZnTPP, red and E1/2 (-0.98 V) is the reduction potential of the acceptor SnTPP.31,37 The excitation energy is ∆E* ) 2.08 eV for ZnTPP in MTHF and 2.09 eV in toluene.32 The average radius of the porphyrin radicals is rd+ ) ra- ) 4.5 Å; the dielectric constants are εS ) 7.52 and 2.4 for THF and toluene, respectively, and εref ) 38.25 for DMF. The values of ∆G° for the compounds T-n in THF and toluene are given in Table 4, together with the intramolecular distances R. The free energy changes for oxidative and reductive quenching are nearly identical, as the excitation energy is nearly the same for ZnTPP and SnTPP. The calculation shows that photoinduced electron transfer is possible for the compounds in THF solution. But ∆G° comes out to be positive for toluene as a solvent. Nevertheless, the possibility of electron transfer has to be considered also for this case. As has been discussed earlier,38 the solvation term is not very exact and overestimates the effect of a nonpolar solvent, which leads to wrong predictions for cases of nearly vanishing driving force. For the interpretation of the data obtained with time-resolved fluorescence and transient absorption changes it is best to begin with compound T-2 in toluene. The decay of fluorescence is governed by life times of 220 and 960 ps (Figure 3d). Both of these time constants appear in the TAS measurements: in the repopulation kinetics after light-induced bleaching of the ground state at 562 nm (Figure 5C) and in the buildup of a reaction product, measured at 720 nm (Figure 5D). The agreement of the time constants proves that two photoreactions occur in parallel. Because of the overlapping absorption spectra, both the chromophores are excited. As will be discussed in detail for the results in THF, the electron transfer starting from the first excited singlet state of the zinc porphyrin is the faster reaction. Thus, in spite of the probably very small driving force in toluene, electron-transfer reactions are responsible for the intensity and lifetime quenching of fluorescence. Especially it is plausible that the main decay times correspond to oxidative and reductive quenching of two different excited states. The reactions are slow enough to be detected in our TAS measurements. The radical cation and anion of ZnP and SnP, respectively, are clearly seen in the difference spectra (Figure 4).The back-reaction of the products is too slow as to be followed with our apparatus. For this reason we have made no effort to explain the differing time constants of 53 and 36 ns in the decay of the 720 and 562 nm absorption changes. In THF the driving force for charge separation is much more favorable than in toluene. The two main time constants of roughly 20 and 200 ps in the fluorescence decay show that, as expected, oxidative and reductive quenching are also occurring here. But these reactions are so fast that they cannot always be seen in the TAS measurements. The laser flash used was not short enough. For the formation of the radical pair of T-2 (Figure 5B) we expect a biphasic formation with time constants of 28 and 230 ps. But the fast process occurs concurrently with the

Oligothiophene-Bridged Bisporphyrins

J. Phys. Chem. B, Vol. 113, No. 8, 2009 2533

TABLE 5: Rate Constants for Charge Separation of the Zinc Porphyrin kCS(ox) and of the Tin Porphyrin kCS(red), Calculated with Eq 2, and Constant of Charge Recombination kCR, Calculated Using the Corresponding Lifetimes from the Fits for the TAS Measurements compound

kCS(ox) [s-1]

kCS(red) [s-1]

T-1 T-2 T-3 T-2 toluene

5.8 × 10 3.5 × 1010 3.8 × 1010 4.0 × 109

2.5 × 10 3.0 × 109 4.9 × 109

10

kCR [s-1]

9

6.7 × 109 1.5 × 1010 2.3 × 107

formation of the excited state in the flash, leading to a mixed time of 43 ps. The 230 ps time is not seen as the back-reaction of the radicals (150 ps) is faster than their formation with the consequence of a low stationary radical yield for this process. The time constants of radical formation and decay are better seen in the repopulation kinetics of the ground state (Figure 5A). The ground state is not totally reached within 1 ns. The ratio of persisting and initial ground-state depletion is a measure of the triplet quantum yield (ΦT ) 0.15). This number is an average value for two reactions. Two excited states are formed; appreciable intersystem crossing is, however, only expected for the tin porphyrin because of effective spin-orbit coupling. Nevertheless, it not possible to calculate rate constants of intersystem crossing from the average quantum yield. With T-3 the expected time constants for the radical formation (26 and 160 ps) are not seen for similar reasons as with T-2. We see an increase of the 720 nm absorption (Figure 5F) with a time constant of 42 ps, governed by the flash duration, and a decay with τ3 ) 66 ps, which is faster than for T-2. In the decrease of the ground-state depletion (Figure 5E) we see the two slower time constants (160 and 66 ps), corresponding to the relaxation of the excited singlet state and the back-reaction of the radical pair, respectively. Again some triplet state remains at the end of the experimental time. The calculated triplet yield is 0.1. The next step is to determine the rate constants for charge separation and charge recombination. The constants for the oxidative and reductive quenching reactions can be calculated from the lifetime of fluorescence in the presence (τ) and absence (τ0) of electron transfer.

kCS ) τ-1 - τ0-1

(2)

As the monomeric porphyrin units of our bisporphyrins have not been synthesized, we take the values for τ0 from ZnTPP and SnTPP as an approximation (1.74 and 0.73 ns, respectively). There is a problem to assign our measured lifetimes of about 20 and 200 ps to the zinc and tin porphyrins. We base our assignment on the equality for lifetime and intensity quenching of fluorescence. If we take the average value of 20 and 200 ps from our lifetime measurements, the lifetime quenching would be 0.01 or 0.1 for ZnP and 0.27 or 0.027 for SnP, dependent on the assignment. We deduce the intensity quenching from our analysis of the fluorescence spectra. The ratio of the integrated intensities for T-1 and ZnT-3, for example, is 0.022 and 0.30 for the zinc and tin porphyrin moieties, respectively. We conclude that the short time of about 20 ps corresponds to the oxidative quenching of the ZnP fluorescence, whereas the longer time has to be attributed to the reductive quenching of the SnP fluorescence. Now the corresponding rate constants can be calculated using eq 2. The results are given in Table 5.

Obviously, electron transfer leading to the oxidative quenching of the zinc porphyrin fluorescence is about 1 order of magnitude faster than that one leading to reductive quenching of the tin porphyrin fluorescence. The reason is that different molecular orbitals are involved. In the first reaction the electron starts from the LUMO of the zinc porphyrin, in which there is appreciable electron density on the meso-C atom,14,39 at which the bridge to the tin porphyrin is bound. The conditions for electron transfer are favorable. In the second reaction the electron starts from the HOMO of the zinc porphyrin, in which there is very little electron density on the meso-C atom. In our series T-1 to T-3 the center-to-center distance, determined by the number of thiophene units in the bridge, varies from 12 to 21 Å (Table 4). The distance dependence of electronic coupling and electron transfer40,41 is known to be given by an exponential function

k ) k0 exp[-βR]

(3)

with the damping factor β ranging from 0.8 to 1 Å-1 for saturated spacers42 and from 0.1 to 0.6 Å-1 for conjugated organic bridges.43 Unfortunately, we are not able to verify this dependence. Certainly kCS(ox) is higher for T-1 with the shortest bridge than for T-3. But T-2 behaves irregularly, kCS(ox) being even smaller than for T-3. If we assume the validity of eq 3, we can calculate a value of β ) 0.048 Å-1 from the couple T-1 and T-3. This is a very small value but in accordance with a number for bridges of p-phenylenevinylene oligomers.44 It is interesting to note that compound T-2 does not behave as expected not only regarding the charge separation but also with respect to the shape of the fluorescence spectrum (as discussed above). The differences in geometry, i.e., in the angles between the planes of porphyrin and thiophene as well as in the tilt angles between the transition moments of the two porphyrins (see Table 4), between T-2 and T-1, T-3 are relatively small. Interestingly, they lead to unexpectedly large differences in the properties. It is astonishing that the rate constants of charge separation following the pathway of reductive quenching as well as those of charge recombination do not follow the established dependence on the distance. The largest constants are calculated for the longest bridge. It should be noted that the rate constants of charge recombination follow directly from the TAS measurements and do not depend on any assumption. We can only speculate that the rate-limiting step in both processes is not the electron transfer through the bridge, but the population of the meso-C atom at the edge of the bridge with electrons. In the reductive quenching reaction this population is unfavorable as the HOMO of ZnP is involved. In the charge recombination the electron starts from an orbital which has been the LUMO in the tin porphyrin before reduction. In this case the electron density at the C atom is obviously diminished by the electron attraction of the electropositive central metal. Finally, we discuss the rate constants, obtained for T-2 in toluene. The rate constant of charge separation kCS (oxidative quenching pathway) is only 4 × 109 s-1, i.e., about 9 times smaller than for THF as a solvent. For the charge recombination (kCR ) 2.3 × 107 s-1) the effect of decreasing rate in a nonpolar solvent is much higher (factor 290). The solvent effects known from the literature are clearly smaller. The Wasielewski group32 found factors of 7 and 5 for kCS and kCR, respectively, between the solvents MTHF and toluene in a porphyrin-pyromellitimide dyad, bridged by a phenylene group. For a similar system in

2534 J. Phys. Chem. B, Vol. 113, No. 8, 2009 THF and benzene the Mataga group28 has published factors of 0.87 for kCS and 26 for kCR. For the interpretation of our experiments we have to take into account that the dielectric constant changes both the free energy change and the energy of reorganization. The backreaction in T-2 is so slow for thermodynamic reasons. We have discussed above that the charge separation in toluene has a small driving force, although the exact value cannot be calculated. The charge recombination would then be highly exergonic and most probably in the Marcus inverted region. For the charge separation the reorganization energy is probably determining. In conjugated bridges even with partial coupling as with the bisthiophene in T-2 the influence of the solvent polarity is expected to be larger than for systems with short bridges as cited above. 5. Conclusions As was to be expected from thermodynamic considerations our bisporphyrins T-1, T-2, and T-3 exhibit photoreactivity, in which the tin porphyrin is reduced and the zinc porphyrin oxidized to the corresponding radicals. Although the oligothiophene bridge is not coplanar to the planes of the porphyrin macrocycles, effective electronic coupling of them is accomplished. There is mechanistic and spectral evidence for this coupling. We have shown that the long distance light-induced electron transfer is only slightly dependent on the distance between donor and acceptor. Spectral evidence comes less from absorption, but rather from fluorescence spectra. There is a Stokes shift of up to 50 nm, which is much larger than for similar porphyrins without bridge.34,45 Obviously, electrons are transferred to the bridge in the excited state. The corresponding enlarged transient electrical dipole momentum and the following solvent reorganization lead to an appreciable lowering of energy in case of polar solvent molecules. This is in line with the results of Quintard/Hammarstro¨m.14,15 In their first paper they concluded from the absorption spectra that for bridges of bis- and terthiophene the porphyrin-bridge coupling would be small. In the second paper, however, measurements of emission anisotropy showed that the asymmetric broadening of the Soret band is due to a second transition, aligned along the bridge, which is the expression of significant coupling. An interesting point with our bisporphyrins is that the interpretation of the fluorescence spectra lead us to claim Stokes shifts of different size for the two porphyrin moieties. The relevance of this assumption must be left to further work. Acknowledgment. We thank the Deutsche Forschungsgemeinschaft (DFG) for financial support and Mrs. Andrea Schulz and Gisela Wo¨hlicke for technical assistance. Supporting Information Available: Details of synthesis. This material is available free of charge via the Internet at http:// pubs.acs.org. References and Notes (1) Gust, D.; Moore, Th. A. Science 1989, 244, 35–41. (2) Wasielewski, M. R. Chem. ReV. 1992, 92, 435–461. (3) Schmidt, J. A.; McIntosh, A. R.; Weedon, A. C.; Bolton, J. R.; Connolly, J. S.; Hurley, J. K.; Wasielewski, M. R. J. Am. Chem. Soc. 1988, 110, 1733–1740. (4) Wasielewski, M. R.; Niemczyk, M. P.; Svec, W. A.; Pewitt, E. B. J. Am. Chem. Soc. 1985, 107, 1080–1082. (5) Penfield, K. W.; Miller, J. R.; Paddon-Row, M. N.; Cotsaris, E.; Oliver, A. M.; Hush, N. S. J. Am. Chem. Soc. 1987, 109, 5061–5065. (6) Joran, A. D.; Leland, B. A.; Geller, G. G.; Hopfield, J. J.; Dervan, P. B. J. Am. Chem. Soc. 1984, 106, 6090–6092.

Regehly et al. (7) Finckh, P.; Heitele, H.; Michel-Beyerle, M. E. J. Phys. Chem. 1988, 92, 6584–6590. (8) Miller, J. R.; Beitz, J. J. Chem. Phys. 1981, 74, 6746–6756. (9) Helms, A.; Heiler, D.; McLendon, G. J. Am. Chem. Soc. 1992, 114, 6227–6238. (10) Pettersson, K.; Wiberg, J.; Ljungdahl, Th.; Martensson, J.; Albinsson, B. J. Phys. Chem. A 2006, 110, 319–326. (11) Arnold, D. P.; James, D. A. J. Org. Chem. 1997, 62, 3460–3469. (12) Schwab, P. F. H.; Levin, M. D.; Michl, J. Chem. ReV. 1999, 99, 1863–1934. (13) (a) Wu¨rthner, F.; Vollmer, M. S.; Effenberger, F.; Emele, P.; Meyer, D. U.; Port, H.; Wolf, H. Ch. J. Am. Chem. Soc. 1995, 117, 8090–8099. (b) Vollmer, M. S.; Effenberger, F.; Stu¨mpfig, Th.; Hartschuh, A.; Port, H.; Wolf, H. Ch. J. Org. Chem. 1998, 63, 5080–5087. (14) Odobel, F.; Suresh, S.; Blart, E.; Nicolas, Y.; Quintard, J.-P.; Janvier, P.; Le Questel, J.-Y.; Illien, B.; Rondeau, D.; Richomme, P.; Ha¨upl, T.; Wallin, S.; Hammarstro¨m, L. Chem. Eur. J. 2002, 8, 3027–3046. (15) Kru¨ger, W.; Fuhrhop, J.-H. Angew. Chem., Int. Ed. 1982, 21, 131– 132. (16) Wallin, S.; Hammarstro¨m, L.; Blart, E.; Odobel, F. Photochem. Photobiol. 2006, 5, 828–834. (17) Korth, O.; Hanke, T.; Ru¨ckmann, I.; Ro¨der, B. Exp. Tech. Phys. 1995, 41, 25–36. (18) Makarov, S.; Litwinski, C.; Ermilov, E. A.; Suvorova, O.; Ro¨der, B.; Wo¨hrle, D. Chem. Eur. J. 2006, 12, 1468–1474. (19) Regehly, M.; Ermilov, E. A.; Helmreich, M.; Hirsch, A.; Jux, N.; Ro¨der, B. J. Phys. Chem. B 2007, 111, 998–1006. (20) Material Studio, Packages: Discover and DMOL3, Version 3.2, Accelrys, Inc., San Diego. (21) Diedrich, F.; Stang, P. J. In Metal-Catalyzed Cross-Coupling Reactions; Wiley: Weinheim, 1997.. (22) Rothemund, P. J. Am. Chem. Soc. 1935, 57, 2010–2011. (23) Adler, A. D.; Longo, F. R.; Finarelli, J. D.; Goldmacher, J.; Assour, J.; Korsakoff, L. J. Org. Chem. 1967, 32, 476–476. (24) Youngblood, W. J.; Gryko, D. T.; Lammi, R. K.; Bocian, D. F.; Holten, D.; Lindsey, J. S. J. Org. Chem. 2002, 67, 2111–2117. (25) Fischer, H.; Orth, H. In Die Chemie des Pyrrols; Akademische Verlagsgesellschaft: Leipzig, 1934; Vol. I. (26) Boyle, R. W.; Bruckner, C.; Posakony, J.; James, B. R.; Dolphin, D. Org. Synth. 1999, 76, 287–293. (27) Arsenault, G. P.; Bullock, E.; MacDonald, S. F. J. Am. Chem. Soc. 1960, 82, 4384–4389. (28) Osuka, A.; Nakajima, S.; Maruyama, K.; Mataga, N.; Asahi, T.; Yamazaki, I.; Nishimura, Y.; Ohno, T.; Nozaki, K. J. Am. Chem. Soc. 1993, 115, 4577–4589. (29) Felton, R. H. In The Porphyrins; Dolphin, D., Ed.; Academic Press: New York, 1978; Vol. V, p 84. (30) Fajer, J.; Borg, D. C.; Forman, D.; Dolphin, D.; Felton, R. H. J. Am. Chem. Soc. 1970, 92, 3451–3459. (31) Harriman, A.; Richoux, M. C.; Neta, P. J. Phys. Chem. 1983, 87, 4957–4965. The value of Ered )-0.56 V for Sn(IV)TPPS4-, measured in aqueous solution, can be transformed to-0.75 V in THF. (32) Hayes, R. T.; Walsh, C. H. J.; Wasielewski, M. R. J. Phys. Chem. A 2004, 108, 2375–2381. (33) Wasielewski, M. R.; Johnson, D. G.; Svec, W. A.; Kersey, K. M.; Minsek, D. W. J. Am. Chem. Soc. 1988, 110, 7219–7221. (34) Rimke, I. Diplomarbeit, Humboldt-Universita¨t Berlin, 1996. (35) Ru¨ckmann, I.; Petrouskas, M.; Korth, O.; Hanke, Th.; Ro¨der, B. Opt. Commun. 1997, 134, 379–386. (36) Ru¨ckmann, I.; Zeug, A.; Herter, R.; Ro¨der, B. Photochem. Photobiol. 1997, 66, 576–584. (37) Davis, D. G. In The Porphyrins; Dolphin, D., Ed.; Academic Press: New York, 1978; Vol. V, p 144. The value of Ered )-0.81 V for Sn(IV)TPP(Oac)2 in DMF has been transformed to-0.98 V in THF. (38) Heitele, H.; Po¨llinger, F.; Ha¨berle, T.; Michel-Beyerle, M. E.; Staab, H. A. J. Phys. Chem. 1994, 98, 7402–7410. (39) Longuet-Higgins, H. C.; Rector, C. W.; Platt, I. R. J. Chem. Phys. 1950, 18, 1174–1181. (40) Bixon, M.; Jortner, J. AdV. Chem. Phys. 1999, 106, 35–208. (41) Closs, G.L;.; Miller, J. R. Science 1988, 240, 440–447. (42) Paddon-Row, M. N.; Verhoeven, I. W. New J. Chem. 1991, 15, 107–116. (43) Segal, D.; Nitzan, A.; Davis, W. B.; Wasielewski, M. R.; Ratner, M. A. J. Phys. Chem. B 2000, 104, 3817–3829. (44) Davis, W. B.; Ratner, M. A.; Wasielewski, M. R. J. Am. Chem. Soc. 2001, 123, 7877–7886. (45) Gouterman, M.; Schwarz, F. P.; Smith, P. D. J. Chem. Phys. 1973, 59, 676–690. (46) Bergkamp, M. A.; Dalton, J.; Netzel, T. L. J. Am. Chem. Soc. 1982, 104, 253–259. (47) Kirk, W. J. Phys. Chem. A, in press.

JP808052U