Intramolecular energy transfer dynamics in differently linked zinc porphyrin–dithiaporphyrin dyads

Abstract

Intramolecular energy transfer dynamics in two molecular dyads, in which zinc porphyrin (ZnN₄) and dithiaporphyrin (N₂S₂) units were linked covalently by different bridges, namely phenylene (ph) and diphenylethyne (dpe), were studied employing ultrafast time-resolved transient absorption and fluorescence spectroscopic techniques. The rates of energy transfer in both these dyads are slower than in the corresponding ZnN₄–N₄ dyads, in spite of the better gradient for energy flow in the case of the ZnN₄–N₂S₂ dyads. Quantum chemical calculations reveal that the frontier orbital characteristics of the porphyrins are not significantly altered by sulphur substitution at the acceptor porphyrin core, and thus this does not modify the electronic factor in the energy transfer mechanism. However, a significant decrease in overlap between the absorption spectrum of the donor and the emission spectrum of the acceptor results in lower efficiency of the intramolecular energy transfer. The energy transfer process in dpe-linked dyads follows a through-bond super-exchange mechanism, whereas, in ph-linked dyads, the through-space multipole resonance interaction plays an important role.

---

Introduction

Studies on energy transfer dynamics in multi-chromophoric arrays have long been devoted to mimicking the natural photosynthetic light harvesting mechanism and to developing artificial photosynthetic systems.^1–3 Understanding the rates and mechanisms of energy flow in multi-chromophoric arrays is essential for the rational design of molecular architectures with superior light-harvesting performance. Multi-porphyrin assemblies have attracted special interest due to their role as energy surrogates for natural photosynthetic antenna complexes. A large number of such arrays have been synthesized by covalent as well as noncovalent approaches and excited state properties have been examined with regard to their energy-transfer properties.^4–19 Recently, covalently linked multi-porphyrin arrays have been extensively explored to understand not only the light-harvesting properties but also the energy flow mechanism. To this end, several covalently linked metalloporphyrin–porphyrin dyads with different linkers and linker positions have been systematically synthesized and employed for mechanistic studies of energy transfer.^9–16 In all these studies, free base porphyrin has been used as an energy trap, whereas various metalloporphyrins and linking architectures have been used as light collection and energy donor units. The through-bond Dexter mechanism has been demonstrated as the governing factor determining the energy transfer rate in covalently bonded multi-porphyrin assemblies. Frontier orbital characteristics have been revealed as a crucial factor to decide the rate of energy transfer.^13,14 A common feature among the various multi-porphyrin arrays reported in the literature is that they invariably contain a free base porphyrincore (N₄) as the energy acceptor unit. The energy transfer properties of such symmetrical porphyrin arrays containing N₄ porphyrin cores have been studied by creating an energy gradient between the two or more porphyrin subunits by insertion of a metal such as Zn(II), Mg(II), or Cd(II) in one of the porphyrin subunits and leaving the other porphyrin subunit in the free base form.

Recently, the synthesis of hetero-porphyrin derivatives has been of great interest because of significant variation of the chemical and photophysical properties upon core modification.^20–24 Hetero-porphyrins are prepared by modification of the porphyrin core by replacing one or two nitrogen atoms with another hetero-atom, like oxygen, sulphur etc. These core-modified porphyrins possess a significantly lower singlet state energy than that of the free base porphyrin. This creates a suitable energy gradient in the excited singlet state for efficient energy flow from the metalloporphyrin to the freebase porphyrin. This property can make them suitable for construction of artificial light harvesting systems. A large number of dyads containing hetero-porphyrins have been synthesized to study energy transfer properties.^25–30 Although steady state studies have revealed an efficient energy transfer process, the detailed mechanism of the energy transfer dynamics has not been explored because the energy transfer rates in these dyads have not been reported from time resolved studies having high temporal resolution. In this paper, we describe the energy transfer dynamics in a pair of porphyrin dyads, in which zinc tetraphenylporphyrin (ZnN₄) and dithiaporphyrin (N₂S₂) are linked by two different linker groups, namely, diphenylethynyl and phenyl linkers (Scheme 1). We employed sub-picosecond transient absorption and picosecond fluorescence spectroscopic techniques to determine the energy transfer rates in these two dyads and to compare them with those of the free base porphyrin analogues reported earlier. Differences in the rates of energy transfer between these two classes of porphyrin dyads have been rationalized and mechanistic aspects have been enumerated from comparison of the frontier molecular orbital characteristics of porphyrin and dithiaporphyrin.

Scheme 1 Structures of the two hetero-porphyrindyads and corresponding sub-units used in the present study.

Experimental methods

The detailed synthesis methods of the two hetero-porphyrin dyads, namely, ZnN₄–ph–N₂S₂ (Dyad I) and ZnN₄–dpe–N₂S₂ (Dyad II) as well as the corresponding sub-units, namely, ZnN₄ (III) and N₂S₂ (IV) (Scheme 1), have been reported earlier.^27,28 Toluene of spectroscopic grade (Spectrochem, India) was used as received without further purification in all the spectroscopic measurements. Toluene was chosen because it is a non-coordinating and non-interacting solvent and readily solubilizes porphyrin dyads up to the desired concentrations required for time resolved experiments. Steady state absorption and fluorescence spectra were recorded with a Thermo Spectronic (Biomate-5) absorption spectrometer and a Varian (Cary Eclipse) spectrofluorimeter, respectively. Fluorescence spectra were corrected for the wavelength dependence of the instrument sensitivity.

The dynamics of the excited states were monitored following photoexcitation using 400 nm laser pulses with a femtosecond pump-probe transient absorption spectrometer, which has been described in detail elsewhere.³¹ The pulses of 50 fs duration and 200 μJ per pulse energy at 800 nm at 1 kHz repetition rate were obtained from an amplified laser system (Thales, France). Pump pulses at 400 nm (energy ∼ 5 μJ per pulse) were generated by frequency-doubling of one part of the 800 nm output of the amplifier in a 0.5 mm thick BBO crystal and a small amount from the other part (energy ∼ 1 μJ per pulse) to generate a white light continuum (480–1000 nm) probe in a 2 mm thick sapphire plate. The polarization of the pump beam was fixed at the magic angle with respect to that of the probe beam. The sample solutions were flowed through a quartz cell with a 1 mm path length during measurement. Decay dynamics in a particular wavelength region (10 nm width) was selected using a pair of interference filters placed in front of the photodiodes. The overall time resolution of the absorption spectrometer was about 120 fs. The temporal profiles recorded in the 480–750 nm wavelength region were fitted with up to three exponentially rising or decaying components by an iterative deconvolution method using a sech² type instrument response function with full width at half-maximum of 120 fs, and were also used for constructing the time-resolved differential absorption spectra. All the experiments were performed at 296 K.

Picosecond time resolved fluorescence studies were performed using a streak camera detector (Optoscope-SC10, Optronis, Germany). Samples were excited using 400 nm laser pulses of 50 fs duration obtained from the laser system described above. Fluorescence from the sample was collected (perpendicular to excitation) and focused into the input slit of a spectrograph (Spectral Product, DK240), which spectrally dispersed the fluorescence signal and produced a vertical image at the spectrograph output. The spectrally dispersed fluorescence signal from the spectrograph output was focused into the input slit of the streak camera. Hence a spectrally and temporally resolved fluorescence signal was obtained in a single shot, which was averaged over multiple excitation shots for a better S/N ratio. The temporal resolution of the instrument was measured to be ∼15 ps at a 25 ps mm⁻¹ sweep speed.

Quantum chemical calculations were performed using the GAMESS software package employing the density functional theory (DFT) method using the B3LYP functional and the 6-311 G (d,p) basis set.^32–34

Results and discussion

Steady state study

Fig. 1A shows the absorption spectra of the dyads I and II in toluene. Absorption spectra of the corresponding sub-units III and IV in the same solvent have been shown in Fig. 1B. In each case, the strong absorption band appearing in the 380–460 nm region is assigned to the Soret band and the well-resolved vibronic band appearing in the 500–700 nm region to the Q-band absorption. It is important to note that both the Soret and the Q absorption bands of N₂S₂ are red shifted as compared to those of ZnN₄. This suggests that the energy levels of both the S₂ and S₁ states of N₂S₂ are lower compared to those of the corresponding levels of ZnN₄.

Fig. 1 Steady state absorption spectra of the two dyads (A) and the constituent sub-units (B).

Fig. 1B also shows that the absorption spectrum calculated by addition of those of III and IV agrees well with those of the two dyads. This clearly suggests that the absorption characteristics of both of the sub-units III and IV are not significantly altered by integrating them into the dyads and hence indicates a very weak electronic interaction between the two subunits as well as with the linker in the ground state of each of the two dyads.

The emission spectra of the dyads I and II are shown in Fig. 2A and B, respectively. For each of these two dyads, the emission spectrum recorded using 415 nm photoexcitation consists of dual emission bands. The relatively weaker one, appearing in the 470–670 nm region, consists of two vibronic bands, which are characteristic of fluorescence emission originating from the S₁ state of the sub-unit ZnN₄ (III) (not shown in Fig. 2, vide infra). The more intense emission band appearing in the 670–850 nm region with a maximum at ca. 720 nm is characteristic of the sub-unit N₂S₂ (IV). However, upon photoexcitation at 440 nm, which selectively excites the sub-unit N₂S₂, the intensity of the emission from the sub-unit ZnN₄ reduces significantly. The fact that, in spite of 415 nm photoexcitation which selectively excites the ZnN₄ unit, more intense emission has been observed from the N₂S₂ unit suggests that efficient transfer of energy from the S₁ state localized on the ZnN₄ unit to that of the N₂S₂ unit in the dyad takes places. The fluorescence spectra are measured to be independent of concentration (in the range of 10 to 50 μM) and suggest that the observed energy transfer process is intramolecular in nature. In the case of Dyad II, in which two sub-units are linked by a phenyl group, the contribution of the emission from the ZnN₄ unit is negligibly small and hence suggests a very fast and near quantitative transfer of energy from the sub-unit ZnN₄ to the sub-unit N₂S₂. In the case of Dyad I, in which the diphenylethyne group is the linker, the emission spectrum recorded following 415 nm photoexcitation has a small contribution from the ZnN₄ unit and the energy transfer efficiency in this case has been determined to be close to 90%. Thus, steady state studies clearly indicate efficient energy transfer from the ZnN₄ unit to the N₂S₂ unit in the cases of both the dyads. However, it is noted that exclusive excitation to one unit is not possible and thus the energy transfer efficiency calculation from steady state measurement is an approximate one. We would like to mention that the energy transfer rate has been directly measured from time resolved studies and energy transfer efficiencies were determined using experimental energy transfer rates using eqn (1).

Fig. 2 Fluorescence spectra of Dyad I (A) and Dyad II (B) in toluene upon photoexcitation using 415 nm (black line) and 440 nm (red line) light. The dotted blue line shows the fluorescence spectra of the monomeric N₂S₂ unit.

Time resolved fluorescence studies

To investigate the rates and efficiencies of the energy transfer process from the ZnN₄ to the N₂S₂ moiety, we adopted the streak camera based fluorescence spectroscopic technique with about 15 ps time resolution. Time evolution of the transient fluorescence spectra and temporal evolution of the emission intensity monitored at 620 and 720 nm in the case of the Dyad I following photo-excitation at 400 nm are shown in Fig. 3. At 400 nm, we could selectively excite the ZnN₄ unit (see Fig. 1B) to the S₂ state, which quickly relaxes to the S₁ state via internal conversion, and energy transfer to the N₂S₂ unit takes place from the S₁ state of the donor unit. Thus, the energy transfer dynamics could be accurately monitored, even if a small fraction of the dyads were directly excited to the S₂ state, in which the energy is localized on the N₂S₂ unit. The time-resolved emission spectrum (TRES) recorded within the instrument response time (i.e. at 0 ps delay time following photoexcitation) consists of two emission bands with maxima at 620 nm and 720 nm. With an increased delay time, the intensity of the 620 nm band decays and that of the 720 nm band rises concomitantly; analyses of the temporal fluorescence profiles monitored at these two wavelengths reveal that the decay and rise times of the emission intensities, respectively, are nearly equal (16.5 ± 0.5 ps). This suggests that evolution of the TRES presented in Fig. 3 is associated with the process of energy transfer from the ZnN₄ moiety (emission maximum at 620 nm) to the N₂S₂ moiety (emission maximum at 715 nm). Appearance of significant emission intensity at 720 nm even at zero delay time is obviously due to direct excitation of the N₂S₂ unit with 400 nm photoexcitation as well as the limited temporal resolution of the streak camera (∼15 ps). Therefore, a significant amount of energy transfer takes place within the time window of the instrument response time. However, we could determine the rate of energy transfer from the ZnN₄ to N₂S₂ unit from the TRES and the temporal kinetics.

Fig. 3 Time resolved fluorescence spectra of Dyad I in toluene recorded by the streak camera. Inset: temporal profiles (dots) along with the double exponential fit functions (lines), the dotted curve in the inset represents the instrument response function.

The temporal emission profile recorded at 620 nm shows a single exponential decay with a lifetime of 17 ps, followed by another long-lived component, the lifetime of which could not be determined with significant accuracy because of very low intensity as well as the limitation of the streak camera detection recording up to only 1.5 ns. This long-lived emission may be assigned to the ZnN₄ monomer that remains as an impurity,¹⁹ or to the ZnN₄ moiety of those Dyad I molecules which do not have geometrical configurations favourable for energy transfer (vide infra).³¹ The transient profile recorded at 720 nm shows the rise of emission intensity with a lifetime of 16 ps followed by a slow decay of the emission intensity with a lifetime of about 1.2 ns, which agrees well with the S₁ state lifetime of the N₂S₂ molecule reported earlier.²⁵ Thus we determine the energy transfer rate in the case of Dyad I as k_ET (I) = 16.5 ps⁻¹ = 6.06 × 10¹⁰ s⁻¹ and using eqn (1) and the value of k_f (I) = 2.4 ns⁻¹ = 0.04 × 10¹⁰ s⁻¹ the efficiency of the energy transfer process calculated in the case of Dyad I, ξ_ET (I), is 0.98.

We also recorded the TRES and the temporal profiles at 620 and 720 nm using the streak camera in the case of Dyad II in toluene. We observed that the features of the time evolution of the TRES are very similar to those observed in the case of Dyad I, but the time domain, in which the evolution of the TRES takes place, is much longer. The temporal profile recorded at 620 nm reveals the instrument response time-limited rise followed by two exponential decay components. The faster decay component has a lifetime of 60 ps, but the lifetime of the longer lived component could not be determined accurately because of its very small amplitude. The appearance of this component with a smaller amplitude suggests near complete energy transfer from the ZnN₄ moiety to the N₂S₂ moiety. The rise of emission intensity at 720 nm with a lifetime of 60 ps also ensures the occurrence of the energy transfer process, and the rate of this process has been calculated to be k_ET (II) = 1.67 × 10¹⁰ s⁻¹. The efficiency of the energy transfer process ξ_ET (II) calculated using eqn (1) is about 0.96.

Transient absorption studies

The sub-ps time-resolved transient absorption spectroscopic technique, which has better time resolution than that of the streak camera based technique used here, has also been used to study the energy transfer dynamics in the excited states of the dyads. The time evolution of the differential transient absorption spectra recorded following photoexcitation of Dyad I in toluene using 400 nm laser pulses is shown in Fig. 4. We mentioned earlier that 400 nm photons preferentially excite the ZnN₄ moiety. Differential transient absorption spectra consist mainly of several intense excited state absorption (ESA) (positive absorbance) bands throughout the wavelength region (470–750 nm), overlapped by two weak bands with negative absorbance at 530 and 710 nm. The appearance of the well resolved band structures in the differential transient absorption spectra may be attributed to the multiple Q bands in the 500–750 nm region (Fig. 1A) and the strong emission bands at 710 nm (Fig. 2A).

Fig. 4 Time evolution of the transient absorption spectra and temporal profiles recorded at a few selected wavelengths following photoexcitation of Dyad I in toluene using 400 nm laser pulses.

The appearance of two near isosbestic points at ca. 530 and 570 nm suggests that the process of transformation of one excited species into another must be associated with the evolution of the transient spectra. In conjunction with our time resolved fluorescence data, we infer that decreases in absorbance in the 470–530 and 570–770 nm regions are associated with the decay of the S₁ state, in which the energy is localized on the ZnN₄ moiety, whereas the rise of absorption in the 530–570 nm region represents the energy transfer process from the ZnN₄ moiety to the N₂S₂ moiety. Therefore, the negative absorbance band appearing at 530 nm is assigned to bleaching and that at 710 nm to stimulated emission (SE) from the excited state of the dyad, in which the excitation energy is localized on the N₂S₂ moiety.

Temporal profiles of the transient absorption signals recorded at three selected wavelengths along with the multi-exponential fit functions have also been presented in Fig. 4. Each of them could be fitted well with a three component exponential function, where the components had very similar lifetimes. This suggests that each of the transient absorption spectra recorded in the entire 470–770 nm region have contributions from both the excited states, in which the singlet state energy is localized either on ZnN₄ or N₂S₂. The temporal profile recorded at 520 nm is associated with an instrument response time-limited rise of the ESA followed by an ultrafast decay of the ESA with a lifetime of 0.9 ps leading to the bleaching signal, which subsequently rises slowly to a lifetime of 15 ps. The bleaching signal is long-lived (its lifetime is longer than 500 ps). The temporal profile recorded at 710 nm shows the instrument response time-limited rise of the ESA, followed by its double exponential decay with lifetimes of 0.6 and 14 ps leading to a long-lived negative absorbance signal, which may be assigned to stimulated emission from the N₂S₂ moiety. The transient signal monitored at 570 nm also shows the instrument response time-limited rise of the ESA, which initially decays with about a 1.0 ps lifetime but subsequently rises with a lifetime of about 14 ps.

Ultrafast decay of the ESA with an average lifetime of about 0.9 ± 0.2 ps, which is associated with each of the three transient absorption signals, may be assigned to the S₂ state, corresponding to the Soret band. This state is initially populated by absorption of 400 nm light and decays to populate the S₁ state, in which the excitation energy remains stored on the ZnN₄ moiety, via an internal conversion process. The rise of the stimulated emission intensity monitored at 710 nm or the rise of ESA at 520 nm with a lifetime of 14 ps is ascribed to the population of the S₁ state of the dyad, in which the excitation energy is localized on the N₂S₂ moiety, via an energy transfer process. These results and the rate of energy transfer thus calculated (k_ET = 14.5 ps⁻¹ = 6.9 × 10¹⁰ s⁻¹) are in perfect agreement with those obtained in the time-resolved fluorescence experiment.

In the case of Dyad II, time evolution of the transient spectra and the temporal profiles recorded at a few selected wavelengths have been presented in Fig. 5 and are seen to be similar to that of Dyad I. Following the arguments presented above, the ultrafast decay lifetime of ESA (1.2 ± 0.4 ps) is correlated with the internal conversion of the S₂ state to the S₁ state of the dyad, in which the excitation energy remains localized on the ZnN₄ moiety. In addition, the rate of energy transfer from the ZnN₄ moiety to the N₂S₂ moiety (k_ET = 60 ps⁻¹ = 1.7 × 10¹⁰ s⁻¹) in the case of Dyad II could be determined from the rise time of the ground state bleach monitored at 520 nm, the SE at 710 nm or the ESA at 570 nm.

Fig. 5 Time evolution of the transient absorption spectra and temporal profiles recorded at a few selected wavelengths following photoexcitation of Dyad II in toluene using 400 nm laser pulses.

Mechanism of energy transfer

The time-resolved transient absorption and fluorescence spectroscopic studies presented in the earlier sections reveal that the rate of energy transfer in the case of Dyad II is about four times slower than that in the case of Dyad I. Since both dyads have the same donor and acceptor moieties, the difference in the energy transfer rates can obviously be attributed to the differences in the characteristics of the spacers connecting them. While the structures of the spacers in these two dyads are very similar, the faster energy transfer rate in the case of Dyad I can be assigned to the smaller distance between the donor and the acceptor.

We compare the energy transfer rates determined for the dyads in this work (N₂S₂ substituted dyads) with those already reported for two N₄ substituted dyads, in which the donor (ZnN₄) and the spacers are same.^15,16 We find that the energy transfer rates in the N₂S₂ substituted dyads are nearly four times (for ph linked) and 2.5 times (for dpe linked) slower as compared to the corresponding N₄ substituted dyads. Therefore, although the N₂S₂ substituted dyads provide a better energy gradient for energy flow due to the lower energy of the S₁ electronic state of N₂S₂, the energy transfer process in the hetero-porphyrin system is less efficient compared to that in the N₄ analogue. A series of papers published by Lindsey and coworkers has established the fact that the frontier orbital compositions of the energy donor and acceptor porphyrin units greatly influence the through-bond energy transfer dynamics.^12–15 In the case of ZnN₄, the a_2u orbital acts as the HOMO, possessing a high electron density at the meso position and a low electron density at the β position.¹³ Thus, in the meso-substituted ZnN₄–N₄ dyad, the energy transfer rate is found to be remarkably faster than the β-linked dyad due to the stronger through-bond electronic interaction in the meso-linked derivative.

In contrast, following pentafluoro substitution of the four phenyl rings of the ZnN₄ unit, a reversal of orbital ordering occurs leading to a reversal in the energy transfer rate in the meso- and β-linked dyads.¹⁴ In the present study, the donor unit and the linker position for both the dyads are the same as Lindsey’s dyads, whereas the acceptor unit is different, and thus any difference in electronic interaction may originate from the acceptor site. Hence, it is necessary to know the effect of core modification by sulphur substitution on the frontier orbital characteristics, which participate in the electronic interaction and may significantly alter the energy transfer rate. Fig. 6 shows the DFT calculated frontier molecular orbitals of porphyrin (N₄) and dithiaporphyrin (N₂S₂). It is apparent from the frontier orbital pictures that both of the N₄ and N₂S₂ porphyrins have nearly similar orbital characteristics, which suggests that core modification upon sulphur substitution does not influence the HOMO–LUMO composition. This clearly indicates that the through-bond electronic interaction is essentially similar in the N₄ and N₂S₂ systems. Thus, in comparison to the N₄-porphyrin, the N₂S₂-porphyrin is not expected to influence the energy transfer dynamics via electronic interaction.

Fig. 6 Frontier molecular orbitals of the N₄ and N₂S₂ porphyrins, which influence energy transfer dynamics. Hydrogen atoms are omitted in the figures.

Numerous studies on intramolecularly connected energy donor and acceptor systems reveal that the electronic coupling for the energy transfer process follows one or more of three possible mechanisms: (a) the Förster mechanism of dipole–dipole resonant (coulombic) or through-space interaction, which is active at longer distances up to 100 Å;³⁵ (b) the Dexter mechanism of the electron exchange interaction, which requires an orbital overlap and is active at distances less than 10 Å;³⁶ and (c) super-exchange electronic coupling between the donor, bridge and the acceptor.³⁷ In the first two cases, the bridge is considered as an inert spacer. The super-exchange coupling is believed to decay exponentially with distance (vide infra).

For the energy transfer process of the Förster mechanism, along with other factors, the rate of energy transfer is dependent upon the degree of overlap between the donor emission and the acceptor absorption spectra. The spectral overlap integral for the resonance interaction, J_F (in M⁻¹ cm³), is determined by eqn (2).

The rate of energy transfer (in s⁻¹) by the Förster mechanism can be calculated using eqn (3)

where Φ_D and τ_D are the fluorescence quantum yield and the lifetime of the donor alone, n is the refractive index of the solvent and r_cc is the donor–acceptor center-to-center distance. On the other hand, the Dexter-type energy transfer rate, k_D, is given by eqn (4),where J_D (in cm) is the exchange overlap integral and can be calculated using eqn (5) and V is the electronic coupling matrix element.

Thus, the overlap between the donor emission and acceptor absorption spectra is an important parameter in both the Förster and Dexter mechanisms because the values of both the resonance integral and the exchange integral depend on the magnitude of the overlap integral. The donor emission and acceptor absorption spectra are shown in Fig. 7 and the calculated values of the resonance integral and exchange integral are shown in the insets of the figure.

Fig. 7 Comparison of the overlap of the ZnN₄ emission spectrum with the absorption spectra of N₄ (A) and N₂S₂ (B). The calculated values of the resonance integral (J_F, M⁻¹ cm³) and exchange integral (J_D, cm) are given in the insets.

Considering the Förster energy transfer mechanism, employing eqn (2) and (3), the energy transfer rates in the cases of Dyad I and Dyad II have been calculated to be 8.6 × 10⁹ s⁻¹ and 6.5 × 10⁸ s⁻¹, respectively. The following values were used for calculation of the Förster energy transfer rates: Φ_f (ZnN₄) = 0.03, τ (ZnN₄) = 2.4 ns, n (toluene) = 1.49, r = 13 Å (Dyad I) and 20 Å (Dyad II), and k² = 0.84 (a dynamic average of different conformations is assumed due to the fact that the donor and acceptor units can freely rotate around the single bonds).³⁸ These calculated rates are significantly smaller than the experimentally determined rates of energy transfer, which clearly suggests that the dipole–dipole resonance mechanism does not control the energy transfer rate in these dyad systems. Indeed, the dipole–dipole mediated through space contribution (χ_TS) to the experimentally measured energy transfer rate is around 4% and 13% for Dyad I and Dyad II, respectively, whereas, the through-bond contribution (χ_TB) to energy transfer process dominates in both dyads (Table 1). This is in accordance with the energy transfer mechanism observed in numerous meso–meso linked porphyrin dyad assemblies.^12–17 It may be noted that the ratios of the experimentally determined rates to those calculated assuming the Förster mechanism in the dpe linked dyad are significantly larger (about 10–14 times) as compared to those in the case of ph linked dyads (about 4–6 times).

Table 1 Comparison of the rates and efficiencies of energy transfer processes in ZnN₄–N₄ and ZnN₄–N₂S₂ dyads

Dyad	ξET	kET, 10^10 s^−1	JF, 10^−14 cm^3 M^−1	Förster rate 10^9 s^−1	Ratio, JF^a/JF^b	JD 10^−4 cm	Ratio, JD^a/JD^b	χTS	χTB	Ratio, kET^a/kET^b
a Ref. 13 and 15.b This work.
ZnN4–ph–N4^a	0.99	28	6.8	29	3.8	4.35	2.2	0.10	0.90	4.5
ZnN4–ph–N2S2^b	0.99	6.2	1.8	8.6		1.95		0.13	0.87
ZnN4–dpe–N4^a	0.98	4.2	6.8	2.4		4.35		0.05	0.95	2.5
ZnN4–dpe–N2S2^b	0.96	1.7	1.8	0.7		1.95		0.04	0.96

---

On the other hand, the Dexter mechanism via the exchange interaction generally dominates when the donor and acceptor orbitals are in close proximity (within a few Å). In dpe and ph linked porphyrin dyads, the centre-to-centre distance is ∼20 Å and ∼13 Å, respectively. Thus direct orbital overlap between the donor and acceptor units is expected to be too weak to impart Dexter exchange-mediated energy transfer. An alternative mechanism is the bridge mediated through-bond super-exchange interaction, which can provide the necessary electronic coupling for Dexter energy transfer to become feasible. Indeed, through-bond electronic interaction has been established to be the dominant energy transfer pathway in porphyrin dyads with various linker groups.^12–14

In the case of either the Dexter exchange mechanism or the through-bond super-exchange mechanism, the rate of energy transfer should be proportional to the value of the exchange integral (J_D in eqn (5)), provided that the electronic interaction is comparable and that the latter governs the electronic coupling matrix element. The DFT calculated molecular orbital composition (Fig. 6) clearly suggests that the orbital contribution at the meso position of the N₄ and N₂S₂ acceptors is similar. Hence it is reasonable to assume that the electronic coupling matrix element, V, is similar in magnitude for both the N₄ and N₂S₂ acceptor containing dyads. In this situation, the rate of energy transfer is expected to be proportional to J_D. Indeed, the ratio of the values of J_D to the ZnN₄–N₄ pair and to that of the ZnN₄–N₂S₂ pair closely corresponds to the ratio of the energy transfer rates observed in the case of the dpe linked dyads (Table 1). This essentially confirms that the energy transfer rate in Dyad II is controlled by the through-bond super-exchange mechanism.

We calculated the magnitude of the electronic coupling (V) from the experimentally determined values of the energy transfer rates and the exchange integrals using eqn (4). For Dyad II, the value of V is calculated to be 8.58 cm⁻¹. The magnitude of this parameter indicates a moderate coupling between the donor and acceptor components, i.e. that they retain their electronic identity in the dyad form. This is in agreement with the steady state spectroscopic results discussed in Section 3.1. For Dyad I, the value of V is calculated to be 17.5 cm⁻¹. The enhanced value of V in Dyad I is due to the shorter distance between the donor and the acceptor groups as compared to that in Dyad II. The value of the attenuation factor (β) has been calculated to be 0.178 Å⁻¹ using eqn (6).

The value of β ∼ 0.2 is typical for the phenylethyene bridge mediated energy transfer process.³⁸ Hence, the calculation presented above is valid for the super-exchange mediated energy transfer mechanism operative in both the dyads.

However, unlike in the case of the dpe-linked dyads, in the case of the ph-linked dyads, the energy transfer mechanism appears to be more complicated when we compare the energy transfer rate determined in this work with that previously reported for a N₄ substituted dyad. The ratio of the rate of the energy transfer process determined for the N₄-substituted phenyl-linked dyad to that of the N₂S₂-substituted one (i.e. k_ET^a/k_ET^b = 4.5) is higher than the ratio of the calculated values of the exchange integrals (i.e. J_D^a/J_D^b = 2.2). This suggests that the contribution of the energy transfer process occurring via the through-bond mechanism is not the major one in these dyads, unlike in the case of the dpe-linked dyads. In exchange mediated energy transfer, the difference between the energies of the excited state of the donor and the bridge (ΔE_DB) plays an important role. An inverse square dependence of the energy transfer rate on ΔE_DB has been theoretically predicted and experimentally verified.³⁸ Exchange mediated energy transfer in the case of Dyad I is expected to be of lesser importance as ΔE_DB in the case of the ph-bridge is much larger than that in the dpe-bridge case. On the other hand, the purely dipole–dipole mediated Förster resonance energy transfer mechanism cannot account for the measured energy transfer rate, which is about seven times faster than the calculated one (vide supra). Surprisingly, we find that the ratio of the rates of the energy transfer processes in the ph-linked dyads is close to the ratio of the resonance integral (J_F) (Table 1). This possibly suggests that resonance interaction via multipole states may contribute to the energy transfer mechanism. Previous studies have shown that at short distances corrections due to the multipole contribution to the dipole–dipole approximation may increase the energy transfer rate by a factor of five or more.³⁹ In the ph-linked porphyrin dyads, the center-to-center distance between the energy donor and acceptor units is ∼13 Å, which is sufficiently short to impart such an effect. It is also reported by Lindsey et al. that the effect of orbital ordering reversal on the energy transfer rate in the ph-linked dyad is less prominent compared to that of the dpe-linked dyads.¹⁵ This led them to infer that the through-bond mechanism becomes less important in the case of the ph-linked dyad as compared to that of the dpe-linked dyads. The observed rate dependence with linker length or donor/acceptor character could not be fully accounted for by either the pure through-space or the through-bond mechanism. A change from the through-bond to the through-space mechanism is hinted at in previous studies by Lindsey and coworkers.¹⁵ Our present observation is also suggestive of a similarly intriguing mechanistic involvement and warrants further study to understand the through-space versus through-bond energy transfer mechanisms in covalently linked porphyrin dyads.⁴⁰

Conclusions

We have studied the energy transfer dynamics in two molecular dyads, in which zinc porphyrin and dithiaporphyrin subunits are covalently linked by two different molecular bridges, in toluene using ultrafast transient absorption and fluorescence spectroscopic techniques. Experimental results reveal that core modification of the acceptor porphyrin by two sulphur atoms leads to a reduction in the rate of energy transfer, although the efficiency of energy transfer is not significantly affected. Similar to the covalently linked homoporphyrin assembly, the contribution to the total rate of the energy transfer process via the Förster mechanism is nearly one order of magnitude less and the energy transfer dynamics in the dpe-linked dyad is mainly governed by the Dexter exchange mechanism. The significant retardation in the energy transfer rates in the dithiaporphyrin (N₂S₂) based dyads can be quantitatively attributed to the decrease in the exchange integral between the ZnN₄ donor and N₂S₂ acceptor. In ph-linked dyads, the multipole mediated resonance interaction plays a dominant role in the energy transfer dynamics possibly due to the shorter distance between the donor and acceptor units. In both the dyads, the electronic factor, which plays a crucial role in the energy transfer rate of covalently linked porphyrin assemblies, is not influenced by sulphur substitution at the core of porphyrin. This is apparent from the comparison of the DFT calculated HOMO–LUMO characteristics of porphyrin and dithiaporphyrin. Thus, although core modified porphyrins provide a better gradient for energy flow in a multiporphyrin antenna assembly, the decrease in the spectral overlap integral significantly reduces the rate of the energy flow, which may be detrimental to the energy harvesting efficiency.

Notes and references

1. A. K. Burrell, D. L. Officer, P. G. Plieger and D. C. W. Reid, Chem. Rev., 2001, 101, 2751.
2. D. Gust, T. A. Moore and A. L. Moore, Acc. Chem. Res., 1993, 26, 198.
3. V. Balzani, S. Campagna, G. Denti, A. Juris, S. Serroni and M. Venturi, Acc. Chem. Res., 1998, 31, 26.
4. D. Kim and A. Osuka, Acc. Chem. Res., 2004, 37, 735.
5. Y. Nakamura, N. Aratani and A. Osuka, Chem. Soc. Rev., 2007, 36, 831.
6. S. Prathapan, T. E. Johnson and J. S. Lindsey, J. Am. Chem. Soc., 1993, 115, 7519–7520.
7. R. W. Wagner and J. S. Lindsey, J. Am. Chem. Soc., 1994, 116, 9759.
8. R. W. Wagner, J. S. Lindsey, J. Seth, V. Palaniappan and D. F. Bocian, J. Am. Chem. Soc., 1996, 118, 3996–3997.
9. A. A. Bothner-By, J. Dadok, T. E. Johnson and J. S. Lindsey, J. Phys. Chem., 1996, 100, 17551.
10. J. Seth, V. Palaniappan, T. E. Johnson, S. Prathapan, J. S. Lindsey and D. F. Bocian, J. Am. Chem. Soc., 1994, 116, 10578.
11. A. Osuka, N. Tanabe, S. Kawabata, I. Yamazaki and Y. Nishimura, J. Org. Chem., 1995, 60, 7177.
12. D. Holten, D. F. Bocian and J. S. Lindsey, Acc. Chem. Res., 2002, 35, 57.
13. J. P. Strachan, S. Gentemann, J. Seth, W. A. Kalsbeck, J. S. Lindsey, D. Holten and D. F. Bocian, J. Am. Chem. Soc., 1997, 119, 11191.
14. S. I. Yang, J. Seth, T. Balasubramanian, D. Kim, J. S. Lindsey, D. Holten and D. F. Bocian, J. Am. Chem. Soc., 1999, 121, 4008.
15. S. I. Yang, R. K. Lammi, J. Seth, J. A. Riggs, T. Arai, D. Kim, D. F. Bocian, D. Holten and J. S. Lindsey, J. Phys. Chem. B, 1998, 102, 9426.
16. F. Li, S. W. A. Gentemann, J. Seth, J. S. Lindsey, D. Holten and D. F. Bocian, J. Mater. Chem., 1997, 7, 1245.
17. P. Hascoat, S. I. Yang, R. K. Lammi, J. Alley, D. F. Bocian, J. S. Lindsey and D. Holten, Inorg. Chem., 1999, 38, 4849.
18. J. Otsuki, Y. Kanazawa, A. Kaito, D.-M. S. Islam, Y. Araki and O. Ito, Chem.–Eur. J., 2008, 14, 3776.
19. B. Ventura, F. Barigelletti, F. Lodato, D. L. Officer and L. Flamigni, Phys. Chem. Chem. Phys., 2009, 11, 2166.
20. M. J. Broadhurst, R. Grigg and A. W. Johnson, J. Chem. Soc. D, 1969, 1480.
21. M. J. Broadhurst, R. Grigg and A. W. Johnson, J. Chem. Soc. C, 1971, 3681.
22. R. P. Pandian, D. Reddy, N. Chidambaram and T. K. Chandrashekar, Proc. - Indian Acad. Sci., Chem. Sci., 1990, 102, 307.
23. A. Srinivasan, B. Sridevi, M. V. Reddy, S. J. Narayanan and T. K. Chandrashekar, Tetrahedron Lett., 1997, 38, 4149.
24. A. Ulman, J. Manassen, F. Frolow and D. Rabinovich, Inorg. Chem., 1981, 20, 1987.
25. M. Yedukondalu and M. Ravikanth, Coord. Chem. Rev., 2011, 255, 547.
26. I. Gupta and M. Ravikanth, Coord. Chem. Rev., 2006, 250, 468–518.
27. S. Rai and M. Ravikanth, J. Org. Chem., 2008, 73, 8364.
28. M. Yedukondalu and M. Ravikanth, J. Porphyrins Phthalocyanines, 2011, 15, 83.
29. S. Punidha, S. Rai and M. Ravikanth, J. Porphyrins Phthalocyanines, 2008, 12, 1030.
30. S. Rai and M. Ravikanth, Tetrahedron, 2007, 63, 2455.
31. J. A. Mondal, G. Ramakrishna, A. K. Singh, H. N. Ghosh, M. Mariappan, B. G. Maiya, T. Mukherjee and D. K. Palit, J. Phys. Chem., 2004, 108, 7843.
32. A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
33. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1998, 37, 785.
34. M. W. Schmidt, K. K. Baldridge, J. A. Boatz, S. T. Elbert, M. S. Gordon, J. H. Jensen, S. Koseki, N. Matsunaga, K. A. Nguyen, S. J. Su, T. L. Windus, M. Dupuis and J. A. Montgomery, J. Comput. Chem., 1993, 14, 1347.
35. (a) T. Förster, Naturwissenschaften, 1946, 33, 166; (b) T. Förster, Ann. Phys., 1948, 2, 55.
36. L. Dexter, J. Chem. Phys., 1953, 21, 836.
37. H. M. McConnell, J. Chem. Phys., 1961, 35, 508.
38. K. Pettersson, A. Kyrychenko, E. Ronnow, T. Ljungdahl, J. Martensson and B. Albinsson, J. Phys. Chem. A, 2006, 110, 310.
39. J. Chang, J. Chem. Phys., 1967, 77, 3901.
40. R. Metivier, F. Nolde, C. Mullen and T. Basche, Phys. Rev. Lett., 2007, 98, 047802.

---