Physicochemical and Photophysical Studies on Porphyrin-Based

Jul 11, 2007 - D. Amilan Jose,† Atindra D. Shukla,† G. Ramakrishna,‡ Dipak K. Palit,‡. Hirendra N. Ghosh,*,‡ and Amitava Das*,†. Central S...
0 downloads 0 Views 242KB Size
9078

J. Phys. Chem. B 2007, 111, 9078-9087

Physicochemical and Photophysical Studies on Porphyrin-Based Donor-Acceptor Systems: Effect of Redox Potentials on Ultrafast Electron-Transfer Dynamics D. Amilan Jose,† Atindra D. Shukla,† G. Ramakrishna,‡ Dipak K. Palit,‡ Hirendra N. Ghosh,*,‡ and Amitava Das*,† Central Salt and Marine Chemicals Research Institute (CSIR), BhaVnagar 364002, Gujarat, India, and Radiation & Photochemistry DiVision, Bhabha Atomic Research Center, Mumbai, India ReceiVed: January 23, 2007; In Final Form: May 6, 2007

We report new polychromophoric complexes, where different porphyrin (P) derivatives are covalently coupled to a redox active Mo center, MoL*(NO)Cl(X) (L* is the face-capping tridentate ligand tris(3,5-dimethylpyrazolyl) hydroborate and X is a phenoxide/pyridyl/amido derivative of porphyrin). The luminescence quantum yields of the bichromophoric systems (1, 2, and 5) were found to be an order of magnitude less than those of their respective porphyrin precursors. Transient absorption measurements revealed the formation of the porphyrin radical cation species (P•+) and photoinduced electron transfer from the porphyrin moiety to the respective Mo center in 1, 2, and 5. Electrochemical studies showed that the reduction potentials of the acceptor Mo centers in a newly synthesized pyridyl derivative (2; E1/2[MoI/0] ) ∼ -1.4 V vs Ag/AgCl) and previously reported phenoxy- (1; E1/2[MoII/I] ) ∼ -0.3 V vs Ag/AgCl) and amido- (3; E1/2[MoII/I] ) ∼ -0.82 V vs Ag/AgCl) derivatives were varied over a wide range. Thus, studies with these complexes permitted us to correlate the probable effect of this potential gradient on the electron-transfer dynamics. Time-resolved absorption studies, following excitation at the Soret band of the porphyrin fragment in complexes 1, 2, and 5, established that forward electron transfer took place biexponentially from both S2 and S1 states of the porphyrin center to the Mo moiety with time constants 150-250 fs and 8-20 ps, respectively. In the case of MoL*(NO)ClX (where X is pyridine derivative 2), the high reduction potential for the MoI/0 couple allowed electron transfer solely from the S2 state of the porphyrin center. Time constants for the charge recombination process for all complexes were found to be 150-300 ps. Further, electrochemical and EPR studies with the trichromophoric complexes (3 and 4) revealed that the orthogonal orientation of the peripheral phenoxy/ pyridyl rings negated the possibility of any electronic interaction between two paramagnetic Mo centers in the ground state and thereby the spin exchange, which otherwise was observed for related Mo complexes when two Mo centers are separated by a polyene system with comparable or larger separation distances.

Introduction Tetraphenylporphyrin and its various derivatives have received much attention owing to their important role in biologys specifically in photosynthesis.1 In natural photosynthetic systems, the primary electron transfer (ET) step occurs from a porphyrin-based complex.1 This has resulted in innumerable reports on photophysical studies of model compounds in order to gain a better insight into the mechanism that may actually be involved in natural photosynthetic processes.2 The lowest electronic excited state, S1, generally gets populated by excitation in the Q-bands (500-650 nm) of the porphyrin derivatives.3 Most of the reports, available in the literature, involve photoinduced process(es) following electron transfer from the first excited state (S1) of a porphyrin unit, covalently linked to an electron acceptor fragment.2,3 Porphyrin molecules generally show an even more intense absorption band in the region of 400-430 nm, known as the B or Soret band. Although more than one transition is associated with this band, the excited state generated by excitation in this spectral region has been denoted as S2 in the literature.2,3 The large energy gap between S1 and * To whom correspondence should be addressed. E-mail: amitava@ csmcri.org (A.D.); [email protected] (H.N.G.). † Central Salt and Marine Chemicals Research Institute (CSIR). ‡ Bhabha Atomic Research Center.

S2 implies that radiationless deactivation from S2 to the S1 state is slow enough to allow competing processes to occur. Though not many reports are available in the literature2-5 involving studies where the dynamics of the photoinduced electron transfer were studied following excitation of the porphyrin fragment in a donor-acceptor assembly at the Soret band, Anderson et al. have reported electron transfer from higher excited states (S2) in a porphyrin-viologen complex.5 It was argued that coupling between the porphyrin (donor) and viologen (acceptor) was strong enough to allow electron transfer to compete with internal conversion (IC) from the S2 state. However, such reports are rare in the literature.2,5 Earlier, McCleverty and his co-workers had proposed an intramolecular electron-transfer pathway from the S1 state of the porphyrin to the acceptor Mo center following laser flash excitation at 532/620 nm, for bichromophoric complexes where the MoL*(NO)Cl center is covalently linked to the porphyrin unit through peripheral amido and phenoxy functionalities.6 These studies were restricted mostly to the picosecond time domain. The redox potentials for the MoII/I redox couple in amido and phenoxy complexes are known to be ∼ -0.7 and ∼ -0.3 V, respectively.7 MoL*(NO)Cl2 is also known to react with pyridine (py) and its derivatives to form the corresponding pyridyl derivative (MoIL*(NO)Cl(py)), where MoI/0-based

10.1021/jp0705830 CCC: $37.00 © 2007 American Chemical Society Published on Web 07/11/2007

Porphyrin-Based Donor-Acceptor Systems

J. Phys. Chem. B, Vol. 111, No. 30, 2007 9079

SCHEME 1: Molecular Structure Complexes 1-5

SCHEME 2: Schematic Diagram Showing Probable Relaxation Channel and Electron Transfer between Photoexcited Porphyrin (P) Fragment and Acceptor Orbitals for Mo complexes (Potential Values Are Quoted Vs Ag/AgCl)

reduction appears at ∼ -1.4 V.7 These potential values provide an idea about the relative energy levels of the acceptor orbitals in the Mo centers. It is also possible to assess the relative energy levels of the S2 and S1 states of a porphyrin fragment with the help of the steady-state emission spectra and redox potential data reported in the literature for the related systems (Scheme 2).7,8 This reveals that the high reduction potential for the MoI/0 couple (∼ -1.4 V vs Ag/AgCl) in the pyridyl derivative, as in 2, offers the possibility of restricting the electron transfer from the S1 state (Scheme 2), while such a possibility exists for complexes 1 and 5. Thus, synthesis of the pyridine derivative (Scheme 1) not only enabled us to modulate the reduction potential of the acceptor unit (over a range of -0.3 to -1.4 V (vs Ag/AgCl)) and thereby ∆GET in order to study the effect on the dynamics of the possible photoinduced processes; but also it gave us the opportunity to elucidate the associated intramolecular electron-transfer dynamics in the donor-acceptor porphyrin dyad involving the S2 state of the porphyrin fragment. This prompted us to study the electron-transfer dynamics associated with complexes 1, 2, and 5 following excitation at the Soret band of photoactive porphyrin (P) unit (Scheme 2) in the femtosecond time domain. Further, use of a porphyrin derivative as the bridging ligand in linking redox active paramagnetic Mo(I) centers may offer the possibility to study the role of the orthogonal orientation of

the peripheral phenyl ring in the porphyrin fragment in affecting the spin-spin interaction between the two terminal Mo(I) centers.8,9 With this idea in mind, we synthesized two new trichromophoric complexes 3 and 4 (Scheme 1). For related systems, where the relevant orbitals of the two individual redox active metal centers are too far apart to overlap directly, electron-electron or spin-spin exchange may only occur via the participation of the bridging ligand orbitals.7,9,10 Earlier, McCleverty et al. showed that, for binuclear complexes such as [{LCI(NO)MO}2{4,4′-NC5H4-(CHdCH)4-H4C5N}], two paramagnetic NC5H4MoL*(NO)Cl centers separated by ∼20 Å are ferromagnetically coupled with a spin exchange rate of about 108 s-1.9d Studies on these new complexes (3 and 4) and comparison of the data obtained with those reported by McCleverty and his group7,9,11-13 have provided us the appropriate opportunity to examine the role that orthogonal orientation of the peripheral phenyl ring in a porphyrin molecule might play in affecting the ground-state electronic or spin exchange between the two paramagnetic Mo(I) centers. Such examples, where porphyrin derivatives are used as Bridging Ligand (BL) to link two electroactive centers, are scanty in the literature.8 This motivated us to design and synthesize two symmetrical trans-bis-substituted porphyrin derivatives (H2L4 and L5; Scheme 1), which were used in the present study as the BL for synthesizing novel trichromophoric systems (complexes 3 and 4; Scheme 1). Thus, there are two implications arising from the synthesis of these Mo complexes (1-5). The first is that photophysical studies in the ultrafast time domain for complexes 1, 2, and 5 are expected to unravel the dynamics of the photoinduced electron-transfer processes associated with the excitation to the S2 state of the porphyrin core as a function of the calculated thermodynamic feasibility (∆G°ET) for electron transfer (Scheme 2).14 The second is that studies on the complexes 3 and 4 revealed the role of orthogonal orientation of the peripheral pyridine or phenoxy groups, with respect to the planar porphyrin ring, in determining the extent of electronic interaction between the two terminal Mo centers in the ground state.

9080 J. Phys. Chem. B, Vol. 111, No. 30, 2007

Jose et al.

TABLE 1: Spectral Data and Redox Potential Values for the Porphyrin Derivatives and the Respective Mo Complexes spectral data

redox potential data λmax (nm) (CHCl3) (emission)

ligand/complex HL1 L2 H2L3 H2L4 L5 1d 2 3 4 5d

E1/2 (V)b,c (Mo centered) E1/2 (V)b,c 16e/17e 17e/18e (porphyrin-centered)

λmax (nm) (CHCl3) (UV-vis) ( × 104 dm3 mol-1 cm-1)

Q(0,0), Q(0,1)

φfa

419 (155), 482 (sh), 516 (15.6), 550(10.0), 591(7.5), 648 (4.6) 418 (188), 445 (sh), 515 (17.0), 549(9.3), 590 (8.2), 649 (8.8) 419(280), 514(16), 550(8.1), 589(5.2), 643 (3.8) 420 (150), 516 (9.2), 552 (6.16), 592 (4.7), 649 (4.6) 418 (189), 514 (9.0), 550 (3.8), 588 (3.3), 643 (2.2) 246 (2.7), 276 (sh), 288 (1.9), 419 (18.0), 516 (1.9), 550 (1.0), 590 (0.8), 650 (0.9) 242 (2.7), 279 (sh), 285 (1.7), 420 (8.0), 516 (0.8), 552 (0.54), 591 (0.43), 648 (0.39) 240 (20.5), 420 (37.4), 514 (4.8), 556 (3.2), 591 (3.0), 658 (2.5) 240 (21.3), 273 (sh), 283 (5.8), 421 (13.4), 514 (3.2), 557 (3.2) 418 (10.7), 516 (2.7), 558 (1.7), 589 (1.4), 650 (1.1)

652, 709 651, 709 650, 710 652, 709 650, 709 653, 711

0.0414 0.092 0.050 0.0786 0.0702 0.0044

-0.71

650, 708

0.0035

0.17

-1.85

0.43, 0.85, -1.73

648, 712 650, 711 651, 711

0.00134 0.0024 0.003

-0.76 0.19 -1.25

-1.88

0.48, 0.82, -1.71 0.41, 0.86, -1.74 0.48, 0.83, -1.7

0.48, 0.81, -1.72 0.41, 0.84, -1.73 0.43, 0.85, -1.71 0.49, 0.80, -1.72 0.39, 0.85, -1.75 0.46, 0.81, -1.72

a The absolute quantum yield for H2TPP is used as standard for calculating the relative quantum yield data for respective species. b Recorded in CH2Cl2, and potentials are quoted vs the Fc/Fc+ couple. c Redox processes were reversible unless otherwise mentioned, ∆E (Epa - Epc) was within the range 90-120 mV, and potential value reported were for the scan rate 100 mV s-1; ∆E for the Fc/Fc+ couple for an identical experiment was 85 mV. d Data reported here matched well with the earlier reports.6

Experimental Section Materials. MoL*(NO)Cl2,15 MeL1,16 HL1,16 L2,17 H2L3,18 and complexes 16 and 56 were prepared following known methods [tBu4N]-

with necessary modification (Supporting Information). PF6 was used as background electrolyte for electrochemical studies and was recrystallized from ethanolic solution before use. Diisopropylamine, pyridine, and acetonitrile were dried and distilled over CaH2 prior to use. Pyrrole was distilled under reduced pressure before use. All reactions were performed under argon atmosphere, unless stated otherwise. The water used was doubly distilled. All other chemicals and solvents were obtained locally and were used as such without further purification. Physical Measurements. Elemental analyses (C, H, N) were performed using a Perkin-Elmer 4100 elemental analyzer. IR spectra were recorded with a Perkin-Elmer Spectrum GX 2000 spectrometer as KBr pellets. Electronic spectra were recorded on a Shimadzu UV-3101 PC spectrophotometer. 1H NMR spectra were recorded on a Bruker 200 MHz FT NMR (model: Avance-DPX 200). Fast-atom bombardment measurement was carried out on VG-ZAB instruments using 3-nitrobenzyl alcohol as matrix. EPR spectra were recorded on a Bruker ESP-300E spectrometer either at room temperature or at 77 K using a Dewar insert in a sample chamber. Reduction of the MoIIphenoxy center to the corresponding MoI species was performed in situ by adding excess cobaltocene to the sample solution in an EPR tube at room temperature. The paramagnetic species, thus formed, were stable for several hours in the sealed EPR tube. Electrochemical experiments were performed on a CH660A (USA) electrochemical instrument using a conventional three-electrode cell assembly. A saturated Ag/AgCl electrode as reference electrode and platinum as working electrode were used for all measurements. Ferrocene (Fc) was added at the end of each experiment as internal standard. For all measurements, the Fc/Fc+ couple appeared at 0.41 V (vs Ag/AgCl). The potential values quoted in Table 1 are versus the Fc/Fc+ couple. Room-temperature emission spectra and emission quantum yields were obtained using samples in quartz cells 1.00 cm × 1.00 cm in size in a Perkin-Elmer LS 50B luminescence spectrofluorimeter. The emission and excitation slit widths were kept constant (15/15) for all spectra reported here. The fluorescence quantum yields, φf, were estimated (eq 1) in appropriate solvents (as specified) using the integrated emission

intensity of 5,10,15,20-tetraphenylporphyrin (H 2TPP, φf ) 0.11 in benzene) for free base porphyrins as reference.19

φf ) φf′(Isample/Istd)(Astd/Asample)(η2sample/η2std)

(1)

where φf′ is the absolute quantum yield for the porphyrin molecule used as reference, Isample and Istd are the integrated emission intensities, Asample and Astd are the absorbances at the excitation wavelength, and η2sample and η2std are the respective refractive indices. The femtosecond tunable visible spectrometer was developed based on a multipass amplified femtosecond Ti:sapphire laser system from Avesta, Russia (1 kHz repetition rate at 800 nm, 50 fs, 800 µJ/pulse) and described earlier.20,21 The 800 nm output pulse from the multipass amplifier is split into two parts to generate pump and probe pulses. In the present investigation, we have used both 800 nm (fundamental) and its frequency doubled 400 nm as excitation sources. To generate pump pulses at 400 nm, one part of 800 nm with a 200 µJ/pulse is frequency doubled in BBO crystals. To generate visible probe pulses, about 3 µJ of the 800 nm beam is focused onto a 1.5 mm thick sapphire window. The intensity of the 800 nm beam is adjusted by iris size and ND filters to obtain a stable white light continuum in the 400 nm to over 1000 nm region. The probe pulses are split into the signal and reference beams and are detected by two matched photodiodes with variable gain. We have kept the spot sizes of the pump beam and probe beam at the crossing point around 500 and 300 µm, respectively. The excitation energy density (at both 800 and 400 nm) was adjusted to ∼2500 µJ/cm2. The noise level of the white light is about ∼0.5% with occasional spikes due to oscillator fluctuation. We have noticed that most laser noise is low-frequency noise and can be eliminated by comparing the adjacent probe laser pulses (pump blocked vs unblocked using a mechanical chopper). The typical noise in the measured absorbance change is about 400 ps) for all the wavelengths.24 We had attributed the 300 fs time constant to an intramolecular vibrational energy redistribution, the 2 ps constant to vibrational cooling, and the 13 ps constant to thermal equilibration with solvent.24 The longer time component (>400 ps) was assigned to the excited singlet state (S1) lifetime of the porphyrin moiety. To understand the decay time constant of the longer componentswhich was the excited-state lifetime of the porphyrin moietyswe measured the fluorescence lifetime using time-correlated single-photon counting (TCSPC) measurements, and it was found to be 8.8 ( 0.2 ns. To have a better insight into the intramolecular ET dynamics in 1, 2, and 5, we carried out transient absorption measurements in acetonitrile following excitation with a 400 nm laser pulse. First, we would like to discuss the dynamics associated with 2 following photoexcitation in acetonitrile. Transient absorption spectra of 2 at various time delays are shown in Figure 2. The transient spectrum at each time delay consists of positive absorption peaks centered on 470, 570, 610, and 690 nm, a shoulder at 530 nm, and a broad absorption band from 750 to 1000 nm along with bleach at 510 nm. The transient absorption spectra and relative intensities of the corresponding peaks are different compared to those of the pure porphyrin moiety after excitation at 400 nm.24 The relative intensity of the transient absorption peaks after 650 nm was higher in 2, when compared to that of the simple porphyrin moiety. This suggested that on excitation of 2 a new absorption band appeared in the wavelength region 650-950 nm. This broad absorption band was assigned to the porphyrin radical cation (P•+). Assignment of this band was made on the basis of the results obtained in a complementary pulse radiolysis experiment, where TPP•+ was generated selectively by the reaction of a N3• radical with a free 5,10,-

J. Phys. Chem. B, Vol. 111, No. 30, 2007 9083

Figure 2. Transient absorption spectra of 2 in acetonitrile at different delay time (λext ) 400 nm). The transient spectrum at each time delay consists of positive absorption peaks centered on 470 nm, 570 nm, 610 nm, 690 nm, a shoulder at 530 nm with bleach at 510 nm, which was attributed to the photoexcited state of porphyrin moiety. The absorption band in the range 750-1000 nm was attributed to P•+ on top of photoexcited porphyrin. Inset: (A) Kinetic trace at 850 nm in acetonitrile. (B) Optical absorption spectra of 2 in acetonitrile.

15,20-tetraphenylporphyrin (TPP) molecule in a N2O saturated aqueous solution (5% acetonitrile + 95% water).24 Further, an earlier report revealed that P•+ absorbs in the 600-900 nm wavelength region.24 In this context, it was interesting to observe a growth in the 750-1000 nm region within the 1-10 ps time domain in the transient absorption spectra (Figure 2). This suggested that the new transient, which appeared within that time scale, could be attributed to the generation of the corresponding P•+. Thus, electron transfer from the porphyrin to the Mo moiety took place within that time domain. To determine the electron-transfer dynamics, we monitored the kinetics at different wavelengths, and the respective kinetic decay trace at 850 nm for 2 is shown as an inset in Figure 2. The kinetic decay traces could be fitted multiexponentially with different time constants. The kinetics at 710 nm were fitted with the time constants 180 fs (54%), 10 ps (7.2%), and 250-300 ps (31.8%), while the kinetics at 850 nm were fitted to 180 fs (63.5%), 18 ps (-10.7%) > 300 ps (32.8%). It was interesting to see that the decay kinetics for 2 at different wavelengths were very different from those of the pure porphyrin moiety following excitation at 400 nm. It was evident from the kinetic trace at 850 nm that there was a growth until 50 ps, followed by decay up to 460 ps and beyond. We attributed this growth kinetics to forward electron transfer from the porphyrin moiety to the Mo center with a time constant of 15-20 ps. Although we did not observe clear growth kinetics of the cation radical due to the overlap of excited-state absorption with the oxidized state of the porphyrin moiety, a clear signature exists for the formation of the oxidized porphyrin moiety. Then the signal at both the wavelengths decayed with a time constant of 250-300 ps, which we could attribute to the back electron-transfer process from the Mo center to the porphyrin moiety. The presence of the 180 fs component and a residual absorption beyond 480 ps revealed that the excited porphyrin chromophore did not take part in a photoinduced electron-transfer reaction. For some of the photoexcited porphyrin moieties, the S2 energy level relaxed to an S1 energy level and did not take part in electron-transfer processes. We tried to assess the relative energy levels of the S2 and S1 states of the porphyrin fragment and also the energy levels of the acceptor states of the Mo center in 2 with the help of steady-state emission spectra and redox potential data obtained from cyclic voltammetric techniques (Scheme 2). As stated earlier, the oxidation state for the metal center in 2 is +1 formally, and this center is known to undergo a reversible one electron reduction process at -1.4 V (vs Ag/AgCl). On excitation with a 400 nm laser flash, the S2 energy level of the

9084 J. Phys. Chem. B, Vol. 111, No. 30, 2007

Figure 3. Transient absorption spectra of 1 in acetonitrile at different delay times (λext ) 400 nm). The depleted transient spectrum at each time delay consists of positive absorption peaks centered on 470 nm, 530 nm, 570 nm, 620 nm, and 675 nm has been attributed to the photoexcited state of porphyrin moiety. Appreciable bleach at 510 nm indicated the charge separation reaction. The broad absorption band from 700 - 1000 nm was attributed porphyrin cation radical. Inset: Optical absorption spectra of 1 in acetonitrile.

Figure 4. Decay kinetics of the transients of 1 in acetonitrile (a) at 490 nm, and (b) at 900 nm following excitation with 400 nm laser pulse. Inset: decay kinetics in the shorter time scale.

porphyrin moiety was expected to be populated, with such a phenomenon being well documented in the literature.2,3 The relative energy levels, which are expected to be associated with an electron-transfer mechanism in 2, are depicted in Scheme 2. This suggests that intramolecular electron transfer could take place from the S2 state of the porphyrin moiety. It is evident from Scheme 2 that the electron-transfer process from the S1 state of the porphyrin moiety to the Mo(I) center in 2 is not thermodynamically viable. However, there is a possibility of internal conversion from the S2 state (Soret band) to the S1 state (Q-band). Kinetic decay traces at 850 (inset of Figure 2) and 710 nm revealed a 180 fs decay component which could be attributed to an intramolecular vibration relaxation. It is very interesting to observe that, at an early point in the experiment, transients at all wavelengths decay very fast and then there is a growth in the 700-1000 nm wavelength region. To determine the intramolecular electron-transfer dynamics, we monitored the kinetics at different wavelengths. Figure 4 shows the kinetic decay traces at 490 and 900 nm, which can be fitted multiexponentially. The kinetics at 490 nm can be fitted with the time constants 250 fs (99.6%), 8 ps (-10.7%), and ∼200 ps (6%); those for 900 nm are 250 fs (56%), 10 ps (-15%), and 150 ps (44%). Once again, the growth kinetics could be attributed to electron transfer from the photoexcited porphyrin moiety to the Mo(II) center in 1, which was around 8-10 ps. Signals at all wavelengths decayed with time constants in the range 150-200 ps and could be attributed to back electron transfer from the reduced Mo center to the P•+. The transient absorption at 490 nm was attributed to the absorption due to the S2 state of the photoexcited porphyrin moiety. So the ultrafast decay component (240 fs) at 490 nm, which decayed almost quantitatively (>99%), gives an idea of how the S2 state either

Jose et al.

Figure 5. Transient absorption spectra of 5 in acetonitrile at different delay time (λext ) 400 nm). The transient spectrum at each time delay consists of positive absorption peaks centered on 470 nm, 530 nm, 570 nm, 620 nm, 675 nm on top of huge bleach peaking at 510 and 560 nm indicated an efficient charge separation reaction. The broad absorption band from 700 - 1000 nm was attributed to charge-separated porphyrin cation radical. Inset: Optical absorption spectra of 5 in acetonitrile.

relaxes to the S1 state or donates an electron to the acceptor Mo(II) center intramolecularly. Again, after 1 ps, the transient absorption at 490 nm shows a growth up to 40 ps with a time constant of ∼8 ps. Now, in complex 1 also, not all the photoexcited S2 state of the porphyrin moiety is involved in intramolecular the electron-transfer reaction to the Mo moiety. An increased absorption yield of P•+ in 1, as compared to that in 2, clearly indicates that the electron-transfer process was more efficient in complex 1. The energy level of the electronaccepting state of the MoII was found to be -0.31 V (vs Ag/ AgCl) from the cyclic voltametric experiments (Table 1). It is interesting to note that the accepting energy level of the Mo center in 1 lies below both the S2 and S1 energies of the porphyrin moiety (Scheme 2). Thus, thermodynamically, electron transfer is possible from both states. As a result, intramolecular electron transfer is more efficient in complex 1 than in complex 2. The two time constants we have observed for the ultrafast decay (∼240 fs) at 490 nm and a growth of 8-10 ps at both 490 and 900 nm could be attributed to electron-transfer time constants from the S2 state and the S1 state to the Mo(II) center in 1, respectively. Finally, we would like to discuss the photoexcited dynamics associated with 5. Transient absorption spectra obtained for 5s following excitation at 400 nmsat various time delays are shown in Figure 5. The transient spectrum at each time delay consists of positive absorption peaks centered on 470, 530, 570, 620, and 675 nm, a huge broad absorption band from 700 to 1000 nm, and a bleach at 510 and 560 nm. These time-resolved spectra looked very different when compared to those for complexes 1 (Figure 2) and 2 (Figure 3). Transient spectra clearly indicated the intramolecular charge separation process for 5 was more efficient than that for 1 and 2. To monitor the charge-transfer dynamics, we analyzed the transient decay traces at different wavelengths. The kinetic decay traces at 510, 570, and 1000 nm are shown in Figure 6. The kinetic data could be fitted multiexponentially and appeared to be very complicated. We have also monitored the growth and decay kinetics at 1000 nm, which could be fitted with 55 fs (61.5%), 20 ps (38.5%), and >220 ps (-100%) time constants. Interestingly, at 1000 nm the ultrafast growth component is ∼50 fs, which is equivalent to the excitation population time of the S2 state. At this wavelength, both excited porphyrin and P•+ have absorption, and this complicated the situation for any meaningful analysis. However, the second growth time constant ∼20 ps was assigned to the electron transfer from the S1 state to the Mo moiety. The

Porphyrin-Based Donor-Acceptor Systems

Figure 6. Decay kinetics of the transients of 5 in acetonitrile (a) at 510 nm, (b) at 570 nm and (c) at 1000 nm after excitation with 400 nm laser pulse. Biexponential growth of the bleach indicates two different electron-transfer time-constants from S2 and S1 states of porphyrin moiety to Mo moiety. Inset: decay kinetics in the shorter time scale.

kinetic data at 510 nm were fitted to time constants of 150 fs (70%), 20 ps (30%), and 220 ps (-100%). The kinetic data at 570 nm were fitted with 55 fs (-69%), 280 fs (87%), 18 ps (13%), and 220 ps (-31%) constants. Interestingly, in both cases, bleach has biexponential growth kinetics and was attributed to the forward electron transfer from two different states of the porphyrin moiety. However, the bleach recovery kinetics could be attributed to the back electron transfer from the reduced Mo center to the P•+. The energy level of the electron accepting state of MoII in 5 was determined to be -0.71 V (Table 1). Thus, the energy levels of the S2 and S1 states of the porphyrin moiety in complex 5 reside well above the energy level of the Mo center (Scheme 2). Thus, electron transfer from both the states is thermodynamically viable. For the kinetic trace at 510 nm, overlap due to the other transients is minimized. Thus, we evaluated the intramolecular electron-transfer time constants from the growth and decay of the bleach. The ultrafast time constant (150 fs) evaluated from the growth of the bleach at 510 nm was attributed to the ultrafast electron-transfer time constant from the S2 state. On the other hand, the 20 ps component is electron transfer from the S1 state to the Mo moiety. Electron-Transfer Dynamics. It is now possible to compare the intramolecular electron-transfer dynamics in all three complexes, 1, 2, and 5. We have determined the kinetics for the forward electron-transfer time constants both from the growth of the P•+ and from the growth of the observed bleach. We have observed in 2 that electron transfer is only possible from the S2 state of the porphyrin moiety. However, in 1 and 5, electron transfer can take place from the S2 and S1 states of the porphyrin. Intramolecular electron-transfer time constants are provided in Table 2 for all three complexes. The groundstate energy levels of the porphyrin moiety (S0) and the electron acceptor Mo centers were determined by cyclic voltammetry (Table 1). The energy levels of the upper excited states (S2 and S1) of the porphyrin moiety have been determined by the combination of excitation and emission spectroscopy and cyclic voltametric techniques and are shown in Table 2. Earlier, Rowley et al. had described the electron-transfer reaction as being mainly from the S1 states of porphyrins to the Mo center in 2 and 5, as they could excite only the Q-band (S1, 532 nm) of the porphyrin moiety.6 In the present investigation, we could describe the ET dynamics quantitatively using femtosecond laser spectrometry and could determine these dynamics for electron transfer from the S2 state to the Mo center in all the complexes reported here. We have determined the change of free energy (-∆G0) for the charge separation reaction from the difference of the redox

J. Phys. Chem. B, Vol. 111, No. 30, 2007 9085 potentials between the donor (S2) and acceptor states (Mo) (Scheme 3), and these are shown in Table 2. The electrontransfer time constants from the S2 states to the Mo center have been attributed to τ2, and these are calculated to be 250 fs, 15 ps, and 150 fs for 1, 2, and 5, respectively (Table 2). Both classical27 and quantum mechanical28 theories of electron transfer (ET) predict that ET rates should ultimately decrease with an increase in the thermodynamic driving force (-∆G0).27 We have monitored the ET dynamics for all three complexes in acetonitrile. Further, the structures of the complexes are similar (Scheme 1), and thus, one can imagine that the contribution from the reorganization energy (Λ) (both internal (λi) and solvent (λs)) for the electron-transfer reaction will be similar. In this present investigation, the rate for electron transfer initially increases and eventually decreases with increasing thermodynamic driving force (-∆G0). It is evident from Table 2 that the fastest electron transfer from the S2 state to the Mo moiety took place in 5, where the change in the free energy (-∆G0) of the reaction is 1.16 eV. On the other hand, in 1, the rate for the forward ET was relatively slower (250 fs), where the change in free energy (-∆G0) of the reaction is 1.66 eV. In this investigation, we have observed that the charge separation reaction in all three complexes follows Marcus’s theory of electron-transfer reaction. The change in free energy in charge separation in complex 1 (-∆G0 ) 1.66 eV) is in a higher exoergonic region as compared to the case in 5 (-∆G0 ) 1.17 eV) and results in a slower charge separation time. On the other hand, in 2, the change in free energy is much lower (-∆G0 ) 0.57 eV), which falls in the normal region. In the present investigation, we can see that the electron-transfer reactions from the porphyrin excited states are among the fastest reported in the TPP-Mo complexes as compared to the data reported in the literature for similar kinds of species.23 The question may arise whether this description falls with in the framework of existing electron-transfer theory. In the present investigation, we have observed the highly exothermic (-∆G0 ) 1.66 eV) electron transfer from the S2 state of the porphyrin moiety in 1 and on an ultrafast time scale (250 fs). Ultrafast ET for the inverted region of the electron-transfer reaction has actually put the reaction on the border of the adiabatic limit. However, one may conceive that the reaction from the S2 state proceeds via short-lived excited states of the products, so that the electrontransfer step is nearly activationless.5 A similar explanation was proposed previously for electron transfer from the S1 state in a zinc-porphyrin-quinone system to explain the high rates observed in the inverted region.29 Regarding the charge separation dynamics for all the complexes from the S1 state (Q-band) to the Mo moiety, we have already commented that the ET reaction is not thermodynamically viable from the S1 state in 2, as the acceptor energy level of Mo (+1/0) lies above the S1 state (Scheme 1). However, in two other complexes (1 and 5), ET from the S1 state is feasible. Although we have observed earlier that in these complexes (1 and 5) the majority of the ET reaction takes place from the photoexcited S2 state of the porphyrin moiety and some photoexcited molecules will relax to the S1 state, we have observed biexponential electron-transfer time constants in both 1 and 5, where ultrafast components have been attributed to ET time constants from the S2 state and slower components have been attributed to ET time constants from the S1 state. The thermodynamic driving force (-∆G0) for the charge separation reaction was calculated from the difference of the redox potentials between the donor (S1) and the acceptor states of MoII (Scheme 3) to be 0.71 eV for 1 and 0.22 eV for 5. The

9086 J. Phys. Chem. B, Vol. 111, No. 30, 2007

Jose et al.

TABLE 2: Parameters Relevant to the Electron-Transfer Dynamicsa compd

energy level of the S2 state (V)

energy level of the S1 state (V)

redox potential of the MoI/0 or MoII/I couple (V)

thermodynamic driving force (-∆G°s2/-∆G°s1)

electron-transfer time constant (τ2ET/τ1ET)

1

-1.97

-1.02

-0.31

2 5

-1.99 -1.88

-1.04 -0.93

-1.42 -0.71

-∆G°s2 ) 1.66 eV -∆G°s1 ) 0.71 eV -∆G°s2 ) 0.57 eV -∆G°s2 ) 1.17 eV -∆G°s1 ) 0.22 eV

τ2ΕΤ ) 250 fs τ1ΕΤ ) 8 ps τ2ΕΤ ) 15 ps τ2ΕΤ ) 150 fs τ1ΕΤ ) 20 ps

a The oxidation potentials of the photoexcited S2 and S1 states of the porphyrin moieties of different complexes and the redox potentials of Mo centers are expressed vs Ag/AgCl. -∆G°s2 and -∆G°s1 are the thermodynamic driving forces between donor states (S2 and S1 states of porphyrin) and acceptor states (Mo moiety). τ2ET and τ1ET are the electron-transfer time constants from the S2 and S1 states of the porphyrin moiety to the Mo moiety.

SCHEME 3: Schematic Diagram of Complexes 1, 2 and 5 Showing the Relaxation Channel (Internal Conversion, IC) from S2 and S1 States and Electron Transfer from the S2 (k2ET) and S1 (k1ET) State to Mo center. kr Is the Charge Recombination Time of the Charge Separated Species. All Potential Values Quoted Here Are Vs Ag/AgCl

observed ET time constants for these complexes were found to be 8 and 20 ps, respectively, for electron transfer from the S1 states to Mo(II) centers (Table 2) and were attributed to τ1. The dynamics for electron transfer from the S1 state to the acceptor state of the Mo(II) centers also followed the Marcus theory of ET reaction, which states that the rate of ET increases with an increase in the thermodynamic driving force (-∆G0) and thus falls in the normal region (Table 1). Finally, we have determined the charge recombination for all the above systems by monitoring the kinetics either by monitoring the bleach recovery or the decay of P•+. The charge recombination time in 2 was determined after monitoring the transient at 830 nm, and it was found to be 250-300 ps. For 1, back electron transfer was found to be 150-200 ps after monitoring the decay of the cation radical at 900 nm. Finally, in 5, the charge recombination time, ∼220 ps, was evaluated either by monitoring the bleach kinetics at 490 nm or by following the decay kinetics at 1000 nm. We have determined the change of free energy (-∆G°CR) for the charge recombination reaction from the difference between the redox potentials of the porphyrin and the Mo moiety (Scheme 3) and found it to be 1.21, 2.3, and 1.7 eV for 1, 2, and 5, respectively. Our results are consistent with the rates of charge recombination being in the Marcus inverted region; thus, more exoergonic ∆G° values for charge recombination lead to lower rates. It is interesting to see that although Rowley et al.6 excited the Q-band at 620 nm of molybdenated teraphenylporphyrin, they still found the charge recombination time to be in the range 120-290 ps. The identical rate constants for the charge recombination obtained in the present study with excitation into the S2 state and previous studies with excitation into the S1 state suggest that the electronically and/or vibrationally excited electron-

transfer products of the S2 reaction relax to thermally equilibrated TPP+-(Mo moiety)- on a faster time scale (