tungsten-alkylidyne dyads: generation of reactive porphyrin and metallo radical states.

DOI: 10.1002/chem.201303118

Photoinduced Charge Separation in Zinc–Porphyrin/Tungsten–Alkylidyne Dyads: Generation of Reactive Porphyrin and Metallo Radical States Davis B. Moravec and Michael D. Hopkins*[a] Abstract: The luminescent tungsten–alkylidyne metalloACHTUNGREliACHTUNGREgand [WCl(C-4,4’C6H4CC-py)ACHTUNGRE(dppe)2] (1; dppe = 1,2bis(diphenylphosphino)ethane) and the zinc–tetraarylporphyrins ZnTPP and ZnTPClP (TPP = tetraphenylporphyrin, TPClP = tetra(p-chlorophenyl)porphyACHTUNGRErin) self-assemble in fluorobenzene solution to form the dyads ZnTPP(1) and ZnTPClP(1), in which the metalloACHTUNGREliACHTUNGREgand is axially coordinated to the porphyrin. Excitation of the porphyrin-centered S1 excited states of these dyads initiates intramolecular energy-transfer (ZnPor!1) and electron-transfer (1! ZnPor) processes, which together efficiently quench the S1 state (~ 90 %).

Transient-absorption spectroscopy and an associated kinetic analysis reveal that the net product of the energytransfer process is the 3ACHTUNGRE[dp*] state of coordinated 1, which is formed by S1! 1 ACHTUNGRE[dp*] singlet–singlet (Fçrster) energy transfer followed by 1ACHTUNGRE[dp*]!3ACHTUNGRE[dp*] intersystem crossing. The data also demonstrate that coordinated 1 reductively quenches the porphyrin S1 state to produce the [ZnPor][1+] charge-separated Keywords: carbyne ligands · donor–acceptor systems · electron transfer · photophysics · supramolecular chemistry

Introduction The development of molecular artificial-photosynthetic systems requires the integration of components that harvest light, photoinitiate electron-transfer reactions, and deliver the electrons and holes to catalysts for the reduction and oxidation of solar-fuel feedstocks.[1] Due to the intense interest in this research area there have been numerous studies of homogeneous assemblies that incorporate one or more of these essential components and processes, with the aim of elucidating the geometric, electronic, and thermodynamic design factors necessary to achieve a fully functional artificial-photosynthetic system.[1–4] Because of the steep challenge in integrating fast photophysical and photoredox events with comparatively slow substrate reactions, studies that have focused on probing catalyst photosensitization mechanisms and demonstrating catalytic turnover have typically employed sacrificial reagents as the source of the oxidizing and reducing equivalents, because their prompt decomposition following electron transfer suppresses unproductive charge-recombination pathways. Taking studies of

[a] D. B. Moravec, Prof. M. D. Hopkins Department of Chemistry, The University of Chicago 929 E. 57th Street, Chicago, Illinois (USA) E-mail: [email protected] Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.201303118.

17082

state. This is a rare example of the reductive quenching of zinc porphyrin chromophores. The presence in the [ZnPor][1+] charge-separated states of powerfully reducing zinc–porphyrin radical anions, which are capable of sensitizing a wide range of reductive electrocatalysts, and the 1 + ion, which can initiate the oxidation of H2, proACHTUNGREduces an integrated photochemical system with the thermodynamic capability of driving photoredox processes that result in the transfer of renewable reducing equivalents instead of the consumption of conventional sacrificial donors.

the photochemical reduction of CO2 as a case in point, there have been many reports over the past 30 years[5] in which a mixture of a photosensitizer and a reduction catalyst, or a photosensitizer/catalyst assembly, or an all-in-one photoACHTUNGREcatalyst molecule, has been shown to reduce CO2 to CO or other products using a sacrificial donor, such as a trialkylACHTUNGREamine, as the reductant.[6] By comparison, there are few examples of photochemical assemblies for CO2 reduction that integrate a second catalytic cycle that derives the reducing equivalents from a renewable source, such as H2.[7] Achieving solar-energy storage from homogeneous systems will ultimately require the development of photosensitizers capable of driving these reactions without conventional sacrificial donors. One class of compounds that possesses the properties to fulfill these requirements is that of (dxy)2-configured tungsten–alkylidyne compounds of the type trans-[WACHTUNGRE(CR)L4X] (R = aryl, L = neutral ligand, X = anionic ligand). These luminescent chromophores possess long-lived 3ACHTUNGRE[(dxy)1(p*ACHTUNGRE(WCR))1] excited states (tem = 102–103 ns),[8] are powerful photochemical reductants (E*/ + < 2 V vs. FeCp20/ + ) with small reorganization energies,[9] and their ground-state (dxy)2/ACHTUNGRE(dxy)1 oxidation potentials are systematically tunable across a 2 V range through ligand variation.[10] As a result, these chromophores are well suited for photosensitizing a wide range of electroactive catalysts for difficult-to-reduce substrates, such as CO2. Further, unlike common transitionmetal chromophores, [W(CR)L4X] compounds possess

2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2013, 19, 17082 – 17091

FULL PAPER ground-state reactivity that enables them to derive the reducing equivalents for photoredox reactions from H2 rather than conventional sacrificial donors. Specifically, it has been shown that the d1 oxidation product [WClACHTUNGRE(CPh)ACHTUNGRE(dppe)2] + (dppe = 1,2-bis(diphenylphosphino)ethane) of the representACHTUNGREative chromophore [WClACHTUNGRE(CPh)ACHTUNGRE(dppe)2] reacts in two steps with H2 and a Brønsted base to regenerate the chromophore.[11] The implementation in artificial-photosynthetic reactions of powerful photoredox chromophores that can harvest renewable reducing equivalents sourced to water is an important goal. One way in which [W(CR)L4X] complexes may be implemented in an artificial-photosynthetic system is to assemble them with another redox-active chromophore that is designed to assist in light harvesting and kinetically managing excited-state energy-transfer and electron-transfer processes. In this context, we recently described the structures, bonding, and excited-state properties of self-assembled zinc–porphyrin/tungsten–alkylidyne dyads, in which the luminescent tungsten–alkylidyne metalloACHTUNGREliACHTUNGREgand [WCl(C-4,4’-C6H4CC-py)ACHTUNGRE(dppe)2] (1) is coordinated via its pendant pyridyl moiety to the zinc center of a tetraarylporphyrin (Scheme 1).[12] These

Scheme 1. Zinc–porphyrin/tungsten–alkylidyne ZnTPClP(1).

dyads

ZnTPP(1)

and

dyads, generically denoted hereafter as ZnPor(1), were designed to take advantage of the valuable light-harvesting, photophysical, and redox properties of zinc porphyrins.[13] Density-functional theory calculations on ZnPor(1) showed that the HOMO is the redox-active dxy orbital of 1, and that this and most other frontier orbitals are more than 99 % localized on the parent subunit (Figure 1 a). As a result of this electronic modularity, the porphyrin-centered S1 and T1 excited-state energies are essentially identical to those of simple axial adducts, such as ZnTPP(py), whereas the photophysically important 1ACHTUNGRE[(dxy)1(p*ACHTUNGRE(WCR))1] and 3ACHTUNGRE[(dxy)1(p*ACHTUNGRE(WCR))1] states of 1 (denoted hereafter 1ACHTUNGRE[dp*] and 3ACHTUNGRE[dp*], respectively) are only slightly stabilized (Figure 1 b). This enabled the components of the ZnPor(1) dyads to be de-

Chem. Eur. J. 2013, 19, 17082 – 17091

Figure 1. (a) Orbital energies of ZnTPP(1) and its components (adapted from ref. [12]); the highest-occupied orbitals are indicated. (b) Jablonski diagram of principal photophysical processes (solid arrows) observed for ZnPor(1) in toluene solution; shaded boxes indicate ranges of excitedstate energies for coordinated 1.[14]

signed so that there would be large overlap between the ACHTUNGRE[dxy !p*ACHTUNGRE(WCR)] absorption band of 1 and the S1 fluorescence band of the porphyrin, fulfilling a prerequisite for achievACHTUNGREing intradyad energy transfer. It was found that excitation of the ZnPor(1) S1 state in toluene solution led to highly efficient S1!1ACHTUNGRE[dp*] energy transfer through the Fçrster resonance energy transfer (FRET) mechanism (fEnT ~ 0.8, kEnT ~ 109 s1), with the 1ACHTUNGRE[dp*] state then undergoing fast intersystem crossing (k 1011 s1) to produce the 3ACHTUNGRE[dp*] state (Figure 1 b).[12] Thus, the ZnPor(1) architecture significantly enhances light harvesting for the photochemically useful 3 ACHTUNGRE[dp*] state of 1. Given the electronic modularity of the ZnPor(1) dyads, it seemed reasonable to expect that in polar solvents the redox processes of their porphyrin and metalloACHTUNGREliACHTUNGREgand components would become accessible and that the S1 excited states would be capable of initiating intradyad electron transfer. To explore this possibility, we have now studied the photophysical properties of the ZnPor(1) dyads in the polar solvent fluorobenzene. This solvent was used because it is noncoordinating and has a sufficiently broad electrochemical window to enable study of the relevant photoredox processes. We report here that the S1 excited states of ZnPor(1) dyads are quenched by intradyad electron transfer to form the [ZnPor][1+] charge-separated state, in addition to the FRET pathway observed for the dyads in toluene solution. This represents a rare example in which the zinc porphyrin serves as the primary electron acceptor in a photochemical donor–acceptor system. The [ZnPor][1+] state is a potentially important target for artificial-photosynthesis constructs aimed at achieving CO2 reduction without sacrificial donors, because the ZnPor radical is a powerful reductant[15] capable of reducing all major classes of CO2 reduction electroACHTUNGREcatACHTUNGREalysts to their active redox states,[6, 16] and because the oxidized 1 + subunit possesses the oxidation state necessary for initiating the extraction of the reducing equivalents from H2.[11] 1


www.chemeurj.org

17083

M. D. Hopkins and D. B. Moravec

Results and Discussion

DGCS ¼ e½Eox Ered

Spectroscopic and redox properties of ZnPor(1) dyads in fluorobenzene: In our earlier report,[12] it was shown that the molecular and electronic structures of ZnTPP(1) and ZnTPClP(1) (Scheme 1 and Figure 1) in toluene solution are similar in many respects to those of simple axially substituted ZnPor(L) complexes, such as ZnPor(py), as evidenced by their comparable equilibrium binding constants (ZnTPP(py), Kb = 6000 m1;[17] ZnTPP(1), 7700 m1; 1 ZnTPClP(1); 9100 m ), identical electronic-absorption band maxima, and the axial geometries predicted by density functional theory calculations. In the present study, the axially ligated ZnPor(1) dyads are observed also to form in fluorobenzene. The electronic-absorption spectra observed upon addition of 1 to ZnTPP (Figure 2) and ZnTPClP (Figure S1

e2 E00 4pe0 eS RDA

ð1Þ

where e is the elementary charge, Eox is the oxidation potential of the donor, Ered is the reduction potential of the acceptor, E00 is the excited-state energy, es is the solvent diACHTUNGREelectric constant, and RDA is the donor–acceptor distance. The free energies provided by Equation (1) are set out in Table 1. Unlike the overwhelming majority of porphyrinTable 1. Free energies for photochemical charge separation in ZnPor(1) dyads. DGCS [eV][a] Excited state

ACHTUNGRE[ZnTPP][1+]

ACHTUNGRE[ZnTPClP][1+]

ZnPor S1 ZnPor T1 1 ACHTUNGRE[dp*] 3 ACHTUNGRE[dp*]

0.71 0.24 0.44 to 0.64 0.02 to 0.22

0.79 0.32 0.52 to 0.72 0.12 to 0.32

[a] Calculated from Equation (1) with E00(S1) = 2.04 eV, E00(T1) = 1.57 eV, E00(1ACHTUNGRE[dp*]) = 1.77–1.97 eV,[14] E00(3ACHTUNGRE[dp*]) = 1.35–1.55 eV,[14] es = 5.465,[19] Eox and Ered taken from Table S2 in the Supporting Information, RDA = ACHTUNGRERACHTUNGRE(ZnW) = 15.2 (DFT calculation, ref. [12]).

Figure 2. Electronic-absorption spectra of ZnTPP in fluorobenzene solution upon addition of 1 ([1] = 0–1 mm). Changes to the porphyrin Q bands are indicated by black arrows. Initial isosbestic points are observed at 528 and 555 nm, but are lost with the growth of the 1ACHTUNGRE(dxy !p*) band owing to free 1 (gray arrow).

in the Supporting Information) exhibit porphyrinic QACHTUNGRE(1,0) and QACHTUNGRE(0,0) bands that red shift and change in relative intensity in a manner identical to those of analogous ZnPor(py) complexes in fluorobenzene and of the dyads in toluene solution (Table S1 in the Supporting Information). Based on these close correspondences, it may be concluded that the molecular and electronic structures of the dyads in fluorobenzene and toluene are very similar. The spectra of the dyads do not exhibit the characteristic tungsten–alkylidyne 1 ACHTUNGRE(dxy !p*) and 1ACHTUNGRE(p!p*) bands of 1 because they are much weaker than the porphyrin-centered Soret (e = 105 m1 cm1) and Q bands (e = 104 m1 cm1) that lie in the same regions (for free 1: 1ACHTUNGRE(dxy !p*), lmax = 630 nm, e = 360 m1 cm1; 1ACHTUNGRE(p! p*), lmax = 403 nm, e = 35 000 m1 cm1).[12] Consequently, it is possible to selectively excite the porphyrin-centered S1 states of the dyads. The hypothetical excited-state electron-transfer reactions of the ZnPor(1) dyads are characterized by free-energy changes that can be estimated using Equation (1):[18]

17084

www.chemeurj.org

containing photoredox assemblies, in which S1 excitation results in oxidation of the zinc porphyrin,[3, 4] it is predicted for the ZnPor(1) dyads that production of the [ZnPor][1+] state is downhill from the porphyrin-centered S1 and T1 excited states and from the 3ACHTUNGRE[dp*] and 1ACHTUNGRE[dp*] excited states of coordinated 1. The free energies for charge recombination from the [ZnPor][1+] state to reform the ground state are DGCRACHTUNGRE([ZnTPP][1+]) = 1.33 eV and DGCRACHTUNGRE([ZnTPClP] + [1 ]) = 1.25 eV. The energies of these photoredox states are shown relative to those of the electronic excited states of the dyads in Figure 3. In contrast, the charge-separated

Figure 3. Jablonski diagram of selected states and processes for ZnTPP(1) and ZnTPClP(1) in fluorobenzene. Italicized numbers are energies. Shaded boxes indicate ranges of excited-state energies for coordinated 1.[14] Dashed arrows are hypothetical charge-separation pathways.


Chem. Eur. J. 2013, 19, 17082 – 17091

Zinc–Porphyrin/Tungsten–Alkylidyne Dyads

FULL PAPER

state of form [ZnPor + ][1], which would be produced by oxidative quenching of the porphyrin, is uphill from the S1 excited state by more than 1.2 eV. On the basis of these spectroscopic and energetic considerations, the initial porphyrin-centered S1 excited states of ZnTPP(1) and ZnTPClP(1) produced by selective excitation in the dyad Q bands can decay to the porphyrin T1 state by intersystem crossing, to the 1ACHTUNGRE[dp*] state of coordinated 1 through FRET, and by intradyad electron transfer to form the charge-separated [ZnPor][1+] state. Formation of the [ZnPor][1+] state from the T1, 1ACHTUNGRE[dp*], and 3ACHTUNGRE[dp*] product states of S1 decay is also thermodynamically feasible. Due to the fact that 1ACHTUNGRE[dp*]!3ACHTUNGRE[dp*] intersystem crossing is very fast (kiscW 1011 s1)[12] it is unlikely that formation of [ZnPor] [1+] from the 1ACHTUNGRE[dp*] state is kinetically feasible, and so this pathway is not considered further (or shown in Figure 3).

formation; ZnTPClP(1) tfl = 167 ps, Figure S5 in the Supporting Information), indicating that an additional S1 quenching mechanism is present in fluorobenzene beyond the quenching through FRET observed in toluene. The nature of the processes that quench the S1 excited state of the ZnPor(1) dyads in fluorobenzene solution was probed using transient-absorption spectroscopy. The spectra obtained upon S1 excitation of ZnTPP(1) and ZnTPClP(1) are shown in Figure 4 and Figure 5, respectively. Both sam-

Properties and products of the S1 excited state: As was observed for the ZnPor(1) dyads in toluene solution,[12] the fluorescence of ZnTPP and ZnTPClP in fluorobenzene is significantly quenched upon coordination of 1. The fluorescence decay kinetics of ZnTPP(1) and ZnTPClP(1) observed upon S1 excitation exhibit two distinct processes (Figures S2 and S3 in the Supporting Information). The slower of the two processes is due to fluorescence of unligated ZnPor (ZnTPP, t0 = 2.00 ns; ZnTPClP, 1.13 ns), which is present in small amounts (~ 10–15 %) under the sample conditions, and the faster process is assigned to the fluorescence of the ZnPor(1) dyad (ZnTPP(1), tfl = 263 ps; ZnTPClP(1), tfl < 150 ps; Table 2).[20] For both dyads the fluorescence lifetime is significantly shorter in fluorobenzene solution than in toluene (ZnTPP(1), tfl = 419 ps, Figure S4 in the Supporting In-

Table 2. Photophysical data for ZnPor(1) dyads in fluorobenzene and toluene solution. Parameter

ZnTPP(1) Fluorobenzene Toluene[a]

Fluorescence 263 tfl [ps] 2.00 t0 [ns][b] kq [s1][e] 3.3 109

419 2.1[c] 1.9 109[f]

Transient-absorption spectroscopy 286 2 474 36 tS1 [ps] 3.00 109 1.63 109 kq [s1][h] kEnT [s1] 1.78 109 [i] 1.63 109 kCS [s1][j] 1.22 109 – 0.51 0.77 fEnT[k] 0.35 – fCS[l] 519 4 2200 600 t3ACHTUNGRE[dp*] [ps] not observed – tCR [ps]

ZnTPClP(1) Fluorobenzene Toluene[a] < 150 1.13 > 5.7 109

167 1.0[d] 5.0 109 [g]

123 2 7.25 109 5.68 109[i] 1.57 109 0.70 0.19 695 10 678 11

161 4 5.21 109 5.21 109 – 0.84 – 1261 65 –

[a] Transient-absorption data in toluene from ref. [12]; fluorescence lifetime data from this study; [b] lifetime of free ZnPor; [c] ref. [21]; [d] ref. [22]; [e] kq = 1/tfl1/t0 ; [f] reported as 8.0 108 s1 in ref. [12], as determined by Stern–Volmer analysis of emission-intensity quenching; the emission-lifetime based measurement here is more accurate; [g] reported as 1.9 109 s1 in ref. [12], as determined by Stern–Volmer analysis of emission-intensity quenching; the emission-lifetime based measurement here is more accurate; [h] kq = 1/tS1 1/t0 ; [i] calculated from kEnT measured in toluene[12] using Equation (4); [j] kCS = kqkEnT; [k] fEnT = kEnTtS1; [l] fCS = kCStS1. Chem. Eur. J. 2013, 19, 17082 – 17091

Figure 4. Transient-absorption spectrum of ZnTPP(1) in fluorobenzene (42 % complexed, 68 % selective dyad S1 excitation at lex = 563 nm), showing the appearance (top) and decay (bottom) of S1 product states. Numbers in italics are band maxima [nm]. Values of DA in the 450–700 and 830–1400 nm regions are not comparable because the spectra are acquired using different instruments and laser powers.

ples contain unligated zinc porphyrin, but it is possible to achieve significant selective excitation of the ZnPor(1) dyads in these mixtures because their porphyrin Q bands are red shifted from those of the free porphyrin (Figure 2 and Figure S1 in the Supporting Information). For both dyads, the transient-absorption spectra observed shortly following excitation (Dt = 2.5 ps) exhibit only the features expected for the porphyrin S1 state.[23, 24] These signatures include the strong S1 absorptions at 454 and 1273 nm for ZnTPP(1) and at 465 and 1290 nm for ZnTPClP(1), and negative features in the 600–660 nm region due to QACHTUNGRE(0,0)


www.chemeurj.org

17085


bleaching and QACHTUNGRE(0,0) and QACHTUNGRE(0,1) stimulated emission (ZnTPP(1), 662 nm;[25] ZnTPClP(1), 605 and 655 nm). Over the next few hundred picoseconds the transient-absorption spectra of the two dyads evolve in different ways, and so they will be described individually. For ZnTPP(1) (Figure 4), during the approximately 350 ps following excitation the characteristic porphyrin S1 features at 454, 662, and 1273 nm disappear and an intense band rises at 1110 nm, together with a weak feature at about 600 nm. These new bands were also observed in toluene,[12] and are assigned to the 3ACHTUNGRE[dp*] excited state of coordinated 1. At times longer than about 350 ps the bands at approximately 600 and 1110 nm decay; the weak features that persist at the conclusion of the experiment (Dtffi3 ns; ~ 460 and 850 nm) are due to the porphyrin T1 state. It is not possible to assign these signals individually to ZnTPP(1) or unligated ZnTPP, because their T1 bands are indistinguishable within the resolution of the experiment.[26] The transient-absorption spectrum of ZnTPClP(1) evolves over a period of approximately 230 ps following excitation with loss of the porphyrin S1 features at 605, 655, and 1290 nm, as described above for ZnTPP(1), but the strong absorbance at 465 nm decreases much less in intensity than the corresponding band in ZnTPP(1) and is observed to red shift slightly to 470 nm (Figure 5). Concurrently, three new bands rise in intensity at 615, 912, and 1110 nm. The bands at 615 and 1110 nm are analogous to those seen above for ZnTPP(1) and are assigned to the 3ACHTUNGRE[dp*] excited state of coordinated 1; they were also observed in the transient-absorption spectrum of ZnTPClP(1) in toluene solution (Figure 6).[12] The sharp band at 912 nm was not observed in the spectrum of ZnTPClP(1) in toluene (Figure 6) or in the spectra of ZnTPP(1) in fluorobenzene (Figure 4) or toluene; it is assigned to a porphyrin-centered transition of the charge-separated species [ZnTPClP][1+], based on its close agreement in position and width to a band reported for ZnTPP at 905 nm in DMF (Figure 6; e = 10 100 m1 cm1).[27] The presence of [ZnTPClP][1+] is also indicated by the fact that the transient absorbance near 470 nm is much larger in fluorobenzene than in toluene (Figure 6), due to the contribution from a strong band of ZnTPP in this region (lmax = 455 nm, e = 121 000 m1 cm1 in DMF).[27] The spectrum does not exhibit bands due to 1 + , which is the other component of [ZnTPClP][1+]. However, these bands are expected to be weak relative to porphyrin absorption bands, based on the electronic spectra observed for chemically oxidized [W(CR)L4X] + compounds.[9] At times longer than Dtffi230 ps, the transient-absorption bands of ZnTPClP(1) at 615, 912, and 1110 nm decay and the feature at 470 nm decreases in intensity and shifts to 480 nm (Figure 5). At the conclusion of the experiment (Dtffi3 ns) the spectrum exhibits bands at 480 nm and approximately 850 nm due to the porphyrin T1 state.[23] As for ZnTPP(1), it is not possible to assign the observed T1 signal specifically to ZnTPClP(1) or to unligated ZnTPClP.[26]

17086

www.chemeurj.org

Figure 5. Transient-absorption spectrum of ZnTPClP(1) in fluorobenzene (68 % complexed, 77 % selective dyad S1 excitation at lex = 560 nm), showing the appearance (top) and decay (bottom) of S1 product states. Numbers in italics are band maxima [nm]. Values of DA in the 450–700 and 830–1400 nm regions are not comparable because the spectra are acquired using different instruments and laser powers.

Figure 6. Transient-absorption spectra of ZnTPClP(1) in fluorobenzene (Dt = 233 ps, solid line) and toluene (Dt = 353 ps, dashed line), showing relative intensities at delay times of maximum absorbance of the 1110 nm bands (Figure S6 in the Supporting Information). Features owing to the 3 ACHTUNGRE[dp*] and [ZnTPClP][1+] states are labeled. The absorption band of a chemically prepared sample of ZnTPP (adapted from ref. [27]) is shown by the dotted line.


Chem. Eur. J. 2013, 19, 17082 – 17091


FULL PAPER

Kinetic analysis of S1 excited-state quenching: The fluorescence lifetimes and transient-absorption spectra for ZnTPP(1) and ZnTPClP(1) reveal that there are similarities and differences between the ways their S1 excited states decay relative to each other and as a function of solvent (fluorobenzene and toluene). The transient-absorption spectra of ZnTPP(1) and ZnTPClP(1) in fluorobenzene show that population of their porphyrin-centered S1 excited states results in net S1!3ACHTUNGRE[dp*] energy transfer. This was also observed for the dyads in toluene, and was demonstrated to occur by S1!1ACHTUNGRE[dp*] FRET followed by rapid 1ACHTUNGRE[dp*]!3ACHTUNGRE[dp*] intersystem crossing.[12] Given the close correspondence between the spectroscopic properties of the dyads in fluorobenzene and toluene (vide supra), it is reasonable to conclude that the mechanism of net S1!3ACHTUNGRE[dp*] energy transfer is the same in the two solvents. The spectra for ZnTPClP(1) additionally reveal the formation of the charge-separated [ZnTPClP][1+] state, which is not spectroscopically observed for ZnTPP(1) in fluorobenzene or for either dyad in toluene. As noted above, the charge-separated [ZnTPClP][1+] state lies lower in energy than the porphyrin S1 and T1 states and the 3ACHTUNGRE[dp*] state of 1, and, thus, could be produced from any of them (Figure 3 and Table 1). The mechanism by which the [ZnTPClP][1+] state forms is elucidated by kinetic analysis. The decay kinetics of the S1 excited states of ZnTPP(1) and ZnTPClP(1) are provided by global kinetic analyses of their transient-absorption spectra (see the Supporting Information for details). These analyses (Table 2) provide tS1 = 286(2) ps for ZnTPP(1) (Figure S7 in the Supporting Information), in good agreement with the fluorescence lifetime (tfl = 263 ps), and tS1 = 123(2) ps for ZnTPClP(1) (Figure S8 in the Supporting Information), which is consistent with the upper bound provided by the fluorescence lifetime (tfl < 150 ps). The S1 quenching rate constant (kq) is related to the lifetime of the S1 state with Equation (2): tS1 1 ¼ kS1 ¼ kfl þ kic þ kiscZn þ kq

ð2Þ

where kiscZn is the rate of S1!T1 intersystem crossing, and kfl and kic are the rates for decay to the ground state by fluorescence and internal conversion, respectively. Treating kfl, kic, and kiscZn as independent first-order processes, the rates of which are unchanged from those for free ZnPor (t01 = kfl + kic + kiscZn), provides the S1 quenching rate constants kq = 3.00 109 s1 for ZnTPP(1) and kq = 7.25 109 s1 for ZnTPClP(1) in fluorobenzene (Table 2). The S1 quenching rates for the ZnPor(1) dyads in fluorobenzene solution are over 40 % larger than those observed in toluene (Table 2), in which quenching was established to occur by S1!1ACHTUNGRE[dp*] FRET. It is implausible to account for the accelerated quenching for the dyads in fluorobenzene through an increase in the energy-transfer rate because the physical parameters that govern the rate constant for FRET are largely solvent independent [Eq. (3)]:[28]

Chem. Eur. J. 2013, 19, 17082 – 17091

kFRET ¼

9000lnð10Þk2 FD JF 128p5 h4 NA tD R6DA

ð3Þ

Specifically, the porphyrin radiative lifetime (fD/tD), dyad geometry (which determines RDA, the donor–acceptor center-to-center distance, and k2, the dipole orientation factor), and spectral-overlap integral (JF) are nearly or entirely solvent independent. As a result, the FRET rate constant in fluorobenzene (kFRET,PhF) may be related to that measured in toluene (kFRET,tol) by Equation (4): kFRET;PhF ¼ kFRET;tol

h4tol h4PhF

ð4Þ

where kFRET,tol = 1.63 109 s1 for ZnTPP(1) and 5.21 109 s1 for ZnTPClP(1),[12] htol = 1.494,[19] and hPhF = 1.463.[29] This provides energy-transfer rates for the dyads in fluorobenzene of kEnT = kFRET,PhF = 1.78 109 s1 for ZnTPP(1) and 5.68 109 s1 for ZnTPClP(1) (Table 2). These estimated energy-transfer rates are substantially smaller than the observed quenching rates (ZnTPP(1), kq = 3.00 109 s1; ZnTPClP(1), kq = 7.25 109 s1), indicating the presence of an S1 quenching pathway for both dyads in fluorobenzene solution that is not operative in toluene. For ZnTPClP(1), the nature of this additional process is shown by the transientabsorption spectra to be the reductive quenching of the S1 state by coordinated 1 to form [ZnTPClP][1+] (Figure 5 and Figure 6). For ZnTPP(1), the transient-absorption spectrum does not exhibit corresponding signatures for [ZnTPP][1+], but the thermodynamic driving force for charge separation (DGCS = 0.71 eV, Table 1) is nearly as large as that for [ZnTPClP][1+] (DGCS = 0.79 eV) and the ratio of kEnT to kq is even smaller than that for ZnTPClP(1). Thus, we conclude that the S1 state of ZnTPP(1) is also reductively quenched, but that the relative decay rates of the [ZnTPClP][1+] and [ZnTPP][1+] charge-separated states are such that only the population of the former builds to a spectroscopically detectable amount. Setting kq = kEnT + kCS provides charge-separation rates of kCS = 1.22 109 s1 for ZnTPP(1) and 1.57 109 s1 for ZnTPClP(1); the slightly faster rate for ZnTPClP(1) relative to ZnTPP(1) is in accord with its slightly larger charge-separation driving force (ZnTPP(1), DGCS = 0.71 eV; ZnTPClP(1), DGCS = 0.79 eV; Table 1). The rates and quantum yields for the S1 excited-state decay pathways of the ZnPor(1) dyads in fluorobenzene and toluene are compared in Figure 7. In toluene, the S1 excited state was found to decay principally by S1!1ACHTUNGRE[dp*] FRET (ZnTPP(1), fEnT = 0.77; ZnTPClP(1), fEnT = 0.84), which is followed by fast intersystem crossing on the tungsten–alkylidyne unit to form the 3ACHTUNGRE[dp*] state (kiscW 1011 s1).[12] In fluorobenzene, the quantum yield for FRET is lower than that in toluene (ZnTPP(1), fEnT = 0.51; ZnTPClP(1), fEnT = 0.70) due to the competing charge-separation pathway that forms [ZnPor][1+] (ZnTPP(1), fCS = 0.35; ZnTPClP(1), fCS = 0.19). The porphyrin S1 state also decays through normal porphy-


www.chemeurj.org

17087


so that [ZnTPP][1+] does not build to a spectroscopically detectable concentration. Based on the spectroscopic and kinetic data, it is proposed that the principal charge-recombination pathway for the [ZnPor][1+] state is via the 3 ACHTUNGRE[dp*] state, with which it is in equilibrium, and that the ground state is reformed primarily by nonradiative decay of the 3ACHTUNGRE[dp*] state (Figure 7); charge recombination to directly form the ground state is slow relative to these events. Charge recombination to form the 3 Figure 7. Jablonski diagrams with selected states and processes for ZnTPP(1) (left) and ZnTPClP(1) (right). ACHTUNGRE[dp*] state is energetically [14] Shaded boxes indicate ranges of excited-state energies for coordinated 1. viable because the lower bound for the energy range of the 3 ACHTUNGRE[dp*] state lies only 0.02 eV above the energy of the [ZnTPP][1+] state and 0.12 eV rin-centered intersystem crossing (fiscZn ~ 0.1) and fluorescence (ffl < 0.005). above the [ZnTPClP][1+] state (Table 1). The proposal that the 3ACHTUNGRE[dp*] and [ZnPor][1+] states are in fast equilibrium is Properties and kinetics of the S1 product states: The decay suggested by the fact that for ZnTPClP(1) the 3ACHTUNGRE[dp*] and of the S1 states of ZnTPP(1) and ZnTPClP(1) results in for[ZnTPClP][1+] lifetimes are statistically indistinguishable mation of the ground state and three excited states: the porfrom each other (t3ACHTUNGRE[dp*] = 695ACHTUNGRE(10) and tCR = 678ACHTUNGRE(11) ps) phyrin-centered T1 state, the 3ACHTUNGRE[dp*] state of subunit 1, and and from the results of a kinetic fit that constrains them to the [ZnPor][1+] charge-separated state. The T1 state lies hold the same value (tCR/3[dp*] = 687(8) ps); this is consishighest in energy (Figure 7); therefore, it could decay to the tent with the lifetimes describing the same process. Because 3 the energy barrier between the [ZnTPP][1+] and 3ACHTUNGRE[dp*] ACHTUNGRE[dp*] state through triplet–triplet energy transfer, as observed previously for a covalently linked zinc–porphyrin/ states is nearly zero whereas that for [ZnTPClP][1+] is more [30] tungsten–alkylidyne analogue of ZnPor(1), uphill, charge recombination for [ZnTPP][1+] via this chanand/or by charge separation to the [ZnPor][1+] state, as well as by innel is comparatively efficient and would account for the fact that [ZnTPP][1+] is not detected in the transient-absorption ternal conversion to the ground state. However, the T1 spectrum. The alternative possibility—that the 3ACHTUNGRE[dp*] state states of the dyads are not observed to decay during the time frame of the transient-absorption experiment (Dt = decays via the [ZnPor][1+] charge-separated state, which 3 ns, Figures 4 and 5), and the spectra are complicated by then undergoes charge recombination to form the ground the presence of the T1 bands of free ZnPor. Thus, we do not state—is less consistent with the observations. Although the 3 know the fate of this state and will discuss only the properACHTUNGRE[dp*] lifetimes of the dyads are shorter in fluorobenzene ties of the [ZnPor][1+] and 3ACHTUNGRE[dp*] product states. than in toluene (ZnTPP(1): t3ACHTUNGRE[dp*]ACHTUNGRE(toluene) = 2.20 ns, fluoroGlobal kinetic analysis (see the Supporting Information) benzene = 519 ps; ZnTPClP(1): toluene = 1.26 ns, fluorobenof the transient-absorption spectrum of ZnTPP(1) provides zene = 695 ps; Table 2), the lifetime for ZnTPClP(1) in fluorobenzene is longer than that for ZnTPP(1), and is a larger the 3ACHTUNGRE[dp*] excited-state lifetime t3ACHTUNGRE[dp*] = 519(4) ps (Table 2, percentage of its toluene value; this is inconsistent with Figure S7 in the Supporting Information), and for decay through the [ZnPor][1+] state because it lies more ZnTPClP(1) provides lifetimes for the 3ACHTUNGRE[dp*] and [ZnTPClP] + [1 ] states of t3ACHTUNGRE[dp*] = 695ACHTUNGRE(10) ps and tCR = 678ACHTUNGRE(11) ps, redownhill for ZnTPClP(1) than for ZnTPP(1). The solvent despectively (Table 2, Figure S8 in the Supporting Informapendence of the 3ACHTUNGRE[dp*] lifetime is instead attributed to the tion). The key aspect of the transient-absorption spectra of characteristic nonradiative decay properties of chromophore ZnTPP(1) and ZnTPClP(1) that must be reconciled with 1, based on the fact that the 3ACHTUNGRE[dp*] lifetime of free 1 also exthese kinetic data is that only the spectrum of ZnTPClP(1) hibits a large solvent dependence (toluene, tph = 66 ns;[12] flu + exhibits bands due to the [ZnPor ][1 ] charge-separated orobenzene, 3.9 ns; THF, 4.8 ns). Finally, the fact that the driving forces for charge recombination from the [ZnPor] state, despite the fact that the charge-separation rates for 9 1 the dyads are similar (kCS = 1.22 10 s for ZnTPP(1) and [1+] state to form the ground state are large and similar 9 1 1.57 10 s for ZnTPClP(1)). This requires that charge re(DGCRACHTUNGRE(ZnTPP(1)) = 1.33 eV and DGCRACHTUNGRE(ZnTPClP(1)) = combination for [ZnTPP][1+] be significantly faster than 1.25 eV) is not qualitatively consistent with the significant difference in charge-recombination rates necessary to acthat for [ZnTPClP][1+] (kCRACHTUNGRE(ZnTPP(1)) > kCRACHTUNGRE(ZnTPClP(1)))

17088

www.chemeurj.org


Chem. Eur. J. 2013, 19, 17082 – 17091


count for the observation of [ZnTPClP][1+], but not of [ZnTPP][1+], in the transient-absorption spectra of the dyads.

FULL PAPER tion by 1, that contain the well-studied [Re(CO)3(NN)L] (NN = a,a’-diimine, e.g., bpy) class of CO2 reduction catalysts have been reported by Perutz,[34] Inoue,[35] and Alessio, Iengo, and Indelli.[36] Studies of the photophysics and photochemistry of these triads are underway.

Conclusion The photophysical properties of ZnTPP(1) and ZnTPClP(1) in fluorobenzene solution are both similar and different from those reported previously for the dyads in toluene.[12] In both solvents, fluorescence lifetime measurements and transient-absorption spectroscopy demonstrate that the porphyrin S1 excited state is efficiently (75–90 %) quenched by coordinated 1. In toluene, the sole quenching pathway is S1!1ACHTUNGRE[dp*] energy transfer through the FRET mechanism. In fluorobenzene, the S1 excited state additionally initiates intradyad electron transfer to produce the charge-separated [ZnPor][1+] state. This state is observed by transient-absorption spectroscopy for ZnTPClP(1), and inferred on the basis of kinetic arguments for ZnTPP(1). The fact that [ZnTPClP][1+] is spectroscopically observed but [ZnTPP] [1+] is not, despite their similar driving forces for charge separation from the S1 state and for charge recombination to the ground state, is due to the fact that charge recombination instead produces the nearby 3ACHTUNGRE[dp*] excited state, which is more energetically accessible for [ZnTPP][1+]. There are numerous examples of photoactive zinc–porphyrin-containing assemblies in which the porphyrin S1 excited state initiates an electron transfer reaction. In all but a few of these systems the porphyrin functions as the primary electron donor, yielding a reduced acceptor and oxidized porphyrin.[3, 4] The ZnPor(1) dyads represent rare examples in which the porphyrin serves as the primary electron acceptor. The paucity of such systems is due to the fact that zinc–porphyrins are good photochemical reductants but poor photooxidants in their S1 excited state (ZnTPP*/ + ffi1.6 V, ZnTPP*/ffi0.1 V vs. FeCp20/ + in fluorobenzene). Prior examples of zinc–porphyrin assemblies for which reductive quenching of the S1 state has been reported contain as electron donors ferrocene,[31] carotenoids,[32] and tetrathiafulvalene.[33] In their oxidized forms, these donors do not couple in obvious ways into regenerative catalyst cycles. In contrast, a net photochemical reduction sensitized by the [ZnPor][1+] charge-separated state would leave 1 + as the oxidized donor, which is the active configuration for extracting reducing equivalents from H2. The [ZnPor][1+] charge-separated state is of potential utility in energy storing CO2 reduction reactions. In addition to the reaction chemistry of oxidized tungsten–alkylidyne compounds with H2, the ZnPor radical formed following S1 excitation is a strong reductant (E0/ = 1.75 to 1.95 V vs. FeCp20/ + in polar organic solvents)[15] that is thermodynamically capable of sensitizing members of any common class of CO2 reduction catalysts.[6, 16] This sensitization will be kinetically feasible if the catalyst is appended to the porphyrin periphery to form a catalyst–ZnPor(1) triad. A number of edge-bound porphyrin–catalyst dyads, suitable for coordina-

Chem. Eur. J. 2013, 19, 17082 – 17091

Experimental Section General procedures: All experiments were performed under a nitrogen atmosphere using standard Schlenk and glovebox techniques. Fluorobenzene (Sigma–Aldrich, 99 %) was refluxed over calcium hydride, from which it was transferred under vacuum. Toluene (Burdick & Jackson, high-purity grade) was stored over NaK (1:2) alloy, from which it was transferred under vacuum. The compounds [WCl(C-4,4’-C6H4CCC5H4N)ACHTUNGRE(dppe)2] (1),[12] [WCl(C-4-C6H4CCPh)ACHTUNGRE(dppe)2] (2),[30] and [5,10,15,20tetra(p-chlorophenyl)porphyrinato]zinc (ZnTPClP)[22] were prepared according to standard methods. [5,10,15,20-(Tetraphenyl)porphyrinato]zinc (ZnTPP; Sigma–Aldrich, low-chlorin grade) and pyridine (anhydrous) were used as received. [NnBu4]ACHTUNGRE[PF6] (Fluka, electrochemical grade) was recrystallized from methanol and dried under vacuum at 100 8C for 12 h. Ferrocene was recrystallized from 95 % ethanol and then sublimed under vacuum. Solution samples for optical experiments were prepared on a vacuum line in sealable cuvettes of path-length 1 cm (absorption/emission) or 2 mm (transient absorption), degassed with five freeze–pump– thaw cycles, and sealed under purified nitrogen. Fluorescence lifetime measurements: Fluorescence lifetimes were measured with a ChronosBH time-domain fluorometer (ISS, Inc.) using timecorrelated single photon counting methods. The fluorometer contained Becker–Hickl SPC-130 detection electronics and an HPM-100-40 hybrid PMT detector. Tunable picosecond pulsed excitation (lex = 565 nm) was provided by a Fianium SC400 supercontinuum laser source and integrated acousto-optic tunable filter. Emission wavelengths were selected with bandpass filters (Semrock and Chroma). The instrument response function was measured to be approximately 120 ps FWHM, using a 1 % scattering solution of Ludox LS colloidal silica. Lifetimes were fit though a forward convolution method using the Vinci control and analysis software of the instrument; lifetimes were fit consistently within 5 %. Samples of ZnPor(1) in fluorobenzene and toluene solution were prepared in situ (concentrations: [1]ffi300 mm, [ZnPor]ffi20 mm). The samples were photochemically stable for the duration of the lifetime measurements, as determined by electronic-absorption spectroscopy. Picosecond transient-absorption spectroscopy: Experiments were performed at the Center for Nanoscale Materials, Argonne National Laboratory. The instrument employed a femtosecond Ti:sapphire oscillator regeneratively amplified at 1.7 kHz. The white-light continuum probe was generated using 5 % of the output (sapphire crystal in the visible region, sapphire-containing proprietary material in the near-infrared region). The remaining output was used to drive an optical parametric amplifier that produced tunable femtosecond excitation (pump) pulses. The pump and probe output entered a transient-absorption spectrometer (Helios, Ultrafast Systems), with which the delay of the probe relative to the pump was varied using a mechanical delay line. The pump beam was chopped at one-half the repetition rate of the laser so that an absorption change (DA = ACHTUNGRE[logACHTUNGRE(Ip/I0)]) could be measured as a function of delay, in which Ip and I0 are the intensities of the transmitted probe with the pump on and off, respectively. Spectral content of the probe in the regions 440– 700 nm and 830–1400 nm was collected as a function of delay using separate spectrographs. Second-order diffractions were removed from the near-infrared probe by passing it through an 850 nm long-pass filter before entering the detector. The data were chirp-corrected to within 200 fs over the spectral range by fitting the response of a pure-solvent standard (under identical conditions) to a polynomial function. It was found through experimentation that the TA spectra of ZnPor(1) most amenable to kinetic analysis were of samples in which the extent of ZnPor coordination by 1 was > 40 %; these spectra exhibit signals due to free porphyrin, but not free 1. Excitation into the ZnPor(1) QACHTUNGRE(1,0) ab-


www.chemeurj.org

17089


Acknowledgements This research was supported by the US Department of Energy, Office of Basic Energy Sciences, Solar Photochemistry Program, under Grant DEFG02-07-ER15910. Use of the Center for Nanoscale Materials (CNM) was supported by the US Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC0206CH11357. Fluorescence lifetime measurements were performed in the Institute for Biophysical Dynamics NanoBiology Facility using instrumentation acquired through NIH Grant 1S10RR026988-01. We are grateful to Dr. Gary P. Wiederrecht and Dr. David J. Gosztola of the Center for Nanoscale Materials at Argonne National Laboratory for their assistance with the picosecond transient-absorption measurements, and to Hunter Vibbert for measuring the fluorescence lifetime of 1.

[1] a) J. Barber, P. D. Tran, J. Royal Soc. Interface 2013, 10, 2012098; b) D. Gust, T. A. Moore, A. L. Moore, Faraday Discuss. 2012, 155, 9 – 26; c) B. Rybtchinski, M. R. Wasielewski, in Fundamentals of Materials for Energy and Environmental Sustainability (Eds.: D. S. Ginley, D. Cahen), Cambridge University Press, New York, 2012, pp. 369 – 384; d) W. Song, Z. Chen, M. K. Brennaman, J. J. Concepcion, A. O. T. Patrocinio, N. Y. Marakami Iha, T. J. Meyer, Pure. Appl. Chem. 2011, 83, 749 – 768; e) I. McConnell, G. Li, G. W. Brudvig, Chem. Biol. 2010, 17, 434 – 447; f) A. C. Benniston, A. Harriman, Mater. Today 2008, 11, 26 – 34; g) N. S. Lewis, D. G. Nocera, Proc. Natl. Acad. Sci. USA 2006, 103, 15729 – 15735. [2] a) M. Pter, T. Massimo, in CRC Handbook of Organic Photochemistry and Photobiology, Vol. 1 (Eds.: A. Griesbeck, M. Oelgemçller, F. Ghetti), 3rd ed., CRC, Boca Raton, 2012, pp. 1289 – 1324; b) S. Fukuzumi, K. Ohkubo, J. Mater. Chem. 2012, 22, 4575 – 4587; c) G. Bottari, O. Trukhina, M. Ince, T. Torres, Coord. Chem. Rev. 2012, 256, 2453 – 2477; d) S. Karlsson, J. Boixel, Y. Pellegrin, E. Blart, H. C. Becker, F. Odobel, L. Hammarstrom, Faraday Discuss. 2012, 155, 233 – 252; e) M. R. Wasielewski, Acc. Chem. Res. 2009, 42, 1910 – 1921; f) V. Balzani, A. Credi, M. Venturi, ChemSusChem 2008, 1, 26 – 58; g) H. Imahori, Y. Mori, Y. Matano, J. Photochem. Photobiol. C 2003, 4, 51 – 83; h) D. Gust, T. A. Moore, in The Porphyrin Handbook, Vol. 8 (Eds.: K. M. Kadish, K. M. Smith, R. Guilard), Academic Press, New York, 1999, pp. 153 – 190. [3] Multiporphyrin Arrays (Ed.: D. Kim), Pan Stanford Publishing, Boca Raton, 2012. [4] M. R. Wasielewski, Chem. Rev. 1992, 92, 435 – 461. [5] J. Hawecker, J. M. Lehn, R. Ziessel, J. Chem. Soc. Chem. Commun. 1983, 536 – 538. [6] a) Y. Tamaki, K. Watanabe, K. Koike, H. Inoue, T. Morimoto, O. Ishitani, Faraday Discuss. 2012, 155, 115 – 127; b) B. Kumar, M. Llorente, J. Froehlich, T. Dang, A. Sathrum, C. P. Kubiak, Annu. Rev. Phys. Chem. 2012, 63, 541 – 569; c) C. D. Windle, R. N. Perutz, Coord. Chem. Rev. 2012, 256, 2562 – 2570; d) R. Reithmeier, C. Bruckmeier, B. Rieger, Catalysis 2012, 2, 544 – 571; e) E. B. Cole, A. B. Bocarsly, in Carbon Dioxide as Chemical Feedstock (Ed.: M. Aresta), Wiley-VCH, Weinheim, 2010, pp. 291 – 316; f) A. J. Morris, G. J. Meyer, E. Fujita, Acc. Chem. Res. 2009, 42, 1983 – 1994. [7] J. Ettedgui, Y. Diskin-Posner, L. Weiner, R. Neumann, J. Am. Chem. Soc. 2011, 133, 188 – 190. [8] R. E. Da Re, M. D. Hopkins, Coord. Chem. Rev. 2005, 249, 1396 – 1409.

17090

www.chemeurj.org

[9] B. M. Lovaasen, J. V. Lockard, B. W. Cohen, S. Yang, X. Zhang, C. K. Simpson, L. X. Chen, M. D. Hopkins, Inorg. Chem. 2012, 51, 5660 – 5670. [10] D. E. Haines, D. C. OHanlon, J. Manna, M. K. Jones, S. E. Shaner, J. Sun, M. D. Hopkins, Inorg. Chem. 2013, 52, 9650 – 9658. [11] C. A. Morales-Verdejo, M. I. Newsom, B. W. Cohen, H. B. Vibbert, M. D. Hopkins, Chem. Commun. 2013, 49, 10566 – 10568. [12] D. B. Moravec, M. D. Hopkins, J. Phys. Chem. A 2013, 117, 1744 – 1755. [13] a) L. L. Li, E. W. G. Diau, Chem. Soc. Rev. 2013, 42, 291 – 304; b) Handbook of Porphyrin Science: With Applications to Chemistry, Physics, Materials Science Engineering, Biology and Medicine (Eds.: K. M. Kadish, K. M. Smith, R. Guilard), World Scientific, London, 2010, see especially Vols. 1 and 2; c) M. Aratani, D. Kim, A. Osuka, Acc. Chem. Res. 2009, 42, 1922 – 1934; d) J. Otsuki, J. Porphyrins Phthalocyanines 2009, 13, 1069 – 1081; e) D. Holten, D. F. Bocian, J. S. Lindsey, Acc. Chem. Res. 2002, 35, 57 – 69; f) K. Kalyanasundaram, Photochemistry of Polypyridine and Porphyrin Complexes, Academic Press, London, 1992. [14] The energies of the 1ACHTUNGRE[dp*] and 3ACHTUNGRE[dp*] states are shown as ranges because the 1ACHTUNGRE[dxy !p*ACHTUNGRE(WCR)] and 3ACHTUNGRE[dxy p*ACHTUNGRE(WCR)] transitions are not observed for the dyads. As described in ref. [12], the 0.2 eV range is set by the energies of these transitions for free 1 and the model compound Zn(hexafluoroacetylacetonate)2(1); the energy for ZnPor(1) is expected to lie near the bottom of the range. [15] K. M. Kadish, J. L. Cornillon, C. L. Yao, T. Malinski, G. Gritzner, J. Electroanal. Chem. 1987, 235, 189 – 207. [16] a) C. Costentin, M. Robert, J.-M. Saveant, Chem. Soc. Rev. 2013, 42, 2423 – 2436; b) E. E. Benson, C. P. Kubiak, A. J. Sathrum, J. M. Smieja, Chem. Soc. Rev. 2009, 38, 89 – 99; c) J.-M. Savant, Chem. Rev. 2008, 108, 2348 – 2378. [17] P. Bhyrappa, V. Krishnan, M. Nethaji, J. Chem. Soc. Dalton Trans. 1993, 1901 – 1906. [18] S. E. Braslavsky, Pure Appl. Chem. 2007, 79, 293 – 465. [19] CRC Handbook of Chemistry and Physics (Ed.: W. M. Haynes), 93rd ed., (Internet Version 2013), CRC/Taylor and Francis, Boca Raton, 2013. [20] The fluorescence lifetime of ZnTPClP(1) is close to the temporal response of the instrument (~ 120 ps at FWHM), which prevents an accurate measurement. [21] K. Tomizaki, R. S. Loewe, C. Kirmaier, J. K. Schwartz, J. L. Retsek, D. F. Bocian, D. Holten, J. S. Lindsey, J. Org. Chem. 2002, 67, 6519 – 6534. [22] S. I. Yang, J. Seth, J. P. Strachan, S. Gentemann, D. Kim, D. Holten, J. S. Lindsey, D. F. Bocian, J. Porphyrins Phthalocyanines 1999, 3, 117 – 147. [23] J. Rodriguez, C. Kirmaier, D. Holten, J. Am. Chem. Soc. 1989, 111, 6500 – 6506. [24] D. B. Moravec, B. M. Lovaasen, M. D. Hopkins, J. Photochem. Photobiol. A 2013, 254, 20 – 24. [25] For ZnTPP(1), the QACHTUNGRE(0,0) bleach and overlapping stimulated emission feature are obscured by use of a l 600 nm cut-off filter employed in this experiment to block stray excitation light. [26] V. A. Walters, J. C. de Paula, B. Jackson, C. Nutaitis, K. Hall, J. Lind, K. Cardozo, K. Chandran, D. Raible, C. M. Phillips, J. Phys. Chem. 1995, 99, 1166 – 1171. [27] J. G. Lanese, G. S. Wilson, J. Electrochem. Soc. 1972, 119, 1039 – 1043. [28] a) T. Fçrster, Discuss. Faraday Soc. 1959, 7 – 17; b) D. Guo, T. E. Knight, J. K. McCusker, Science 2011, 334, 1684 – 1687. [29] B. Giner, S. Martn, H. Artigas, M. C. Lpez, C. Lafuente, J. Phys. Chem. B 2006, 110, 17683 – 17690. [30] B. W. Cohen, B. M. Lovaasen, C. K. Simpson, S. D. Cummings, R. F. Dallinger, M. D. Hopkins, Inorg. Chem. 2010, 49, 5777 – 5779. [31] a) H. Imahori, K. Tamaki, D. M. Guldi, C. Luo, M. Fujitsuka, O. Ito, Y. Sakata, S. Fukuzumi, J. Am. Chem. Soc. 2001, 123, 2607 – 2617; b) Y. Kashiwagi, H. Imahori, Y. Araki, O. Ito, K. Yamada, Y. Sakata, S. Fukuzumi, J. Phys. Chem. A 2003, 107, 5515 – 5522; c) F. D’Souza, P. M. Smith, S. Gadde, A. L. McCarty, M. J. Kullman,


!

sorption band resulted in the following relative excited-state populations of dyad and free porphyrin at Dt = 0, as determined from their contributions to the absorption-band profile: ZnTPP(1) 68 %, ZnTPP 32 %; ZnTPClP(1) 77 %, ZnTPClP 23 %. The samples were determined to be photochemically stable for the duration of the experiment by electronicabsorption spectroscopy. Global kinetic analyses of spectra were performed using OriginPro software; full details are provided in the Supporting Information.

Chem. Eur. J. 2013, 19, 17082 – 17091


M. E. Zandler, M. Itou, Y. Araki, O. Ito, J. Phys. Chem. B 2004, 108, 11333 – 11343; d) M. Kubo, Y. Mori, M. Otani, M. Murakami, Y. Ishibashi, M. Yasuda, K. Hosomizu, H. Miyasaka, H. Imahori, S. Nakashima, J. Phys. Chem. A 2007, 111, 5136 – 5143; e) J. Rochford, A. D. Rooney, M. T. Pryce, Inorg. Chem. 2007, 46, 7247 – 7249; f) H. Mansour, M. E. Ei-Khouly, S. Y. Shaban, O. Ito, N. Jux, J. Porphyrins Phthalocyanines 2007, 11, 719 – 728. [32] S. L. Gould, G. Kodis, R. E. Palacios, L. de La Garza, A. Brune, D. Gust, T. A. Moore, A. L. Moore, J. Phys. Chem. B 2004, 108, 10566 – 10580. [33] a) C. K. Gr tzel, M. Gr tzel, J. Phys. Chem. 1982, 86, 2710 – 2714; b) S. Sadaike, K. Takimiya, Y. Aso, T. Otsubo, Tetrahedron Lett. 2003, 44, 161 – 165. [34] a) C. D. Windle, M. V. Campian, A.-K. Duhme-Klair, E. A. Gibson, R. N. Perutz, J. Schneider, Chem. Commun. 2012, 48, 8189 – 8191; b) J. Schneider, K. Q. Vuong, J. A. Calladine, X.-Z. Sun, A. C. Whitwood, M. W. George, R. N. Perutz, Inorg. Chem. 2011, 50, 11877 – 11889; c) A. Gabrielsson, J. R. Lindsay Smith, R. N. Perutz, Dalton Trans. 2008, 4259 – 4269; d) A. Gabrielsson, F. Hartl, H. Zhang, J. R.

Chem. Eur. J. 2013, 19, 17082 – 17091

FULL PAPER Lindsay Smith, M. Towrie, A. Vlcˇek, R. N. Perutz, J. Am. Chem. Soc. 2006, 128, 4253 – 4266; A. Gabrielsson, F. Hartl, J. R. L. Smith, R. N. Perutz, Chem. Commun. 2002, 950 – 951; e) C. J. Aspley, J. R. Lindsay Smith, R. N. Perutz, J. Chem. Soc. Dalton Trans. 1999, 2269 – 2272. [35] a) Y. Kou, S. Nakatani, G. Sunagawa, Y. Tachikawa, D. Masui, T. Shimada, S. Takagi, D. A. Tryk, Y. Nabetani, H. Tachibana, H. Inoue, J. Catal. 2013; DOI: 10.1016/j.jcat.2013.03.025; b) K. Kiyosawa, N. Shiraishi, T. Shimada, D. Masui, H. Tachibana, S. Takagi, O. Ishitani, D. A. Tryk, H. Inoue, J. Phys. Chem. C 2009, 113, 11667 – 11673. [36] a) T. Gatti, P. Cavigli, E. Zangrando, E. Iengo, C. Chiorboli, M. T. Indelli, Inorg. Chem. 2013, 52, 3190 – 3197; b) M. Casanova, E. Zangrando, E. Iengo, E. Alessio, M. T. Indelli, F. Scandola, M. Orlandi, Inorg. Chem. 2008, 47, 10407 – 10418; c) M. Ghirotti, C. Chiorboli, M. T. Indelli, F. Scandola, M. Casanova, E. Iengo, E. Alessio, Inorg. Chim. Acta 2007, 360, 1121 – 1130. Received: August 7, 2013 Published online: November 6, 2013


www.chemeurj.org

17091

The characterization of the saddle shaped nickel(III) porphyrin radical cation: an explicative NMR model for a ferromagnetically coupled metallo-porphyrin radical.

High-potential perfluorinated phthalocyanine-fullerene dyads for generation of high-energy charge-separated states: formation and photoinduced electron-transfer studies.

Communication: Charge-transfer rate constants in zinc-porphyrin-porphyrin-derived dyads: a Fermi golden rule first-principles-based study.

Photoinduced mixed valency in zinc porphyrin dimer of triruthenium cluster dyads.

Time-Resolved Electron Paramagnetic Resonance Study of Photoinduced Electron Transfer in Pd Porphyrin-Quinone and Zn Porphyrin-Quinone Dyads with a Cyclohexylene Spacer.

Metallo-β-Lactamase (MBL)-Producing Enterobacteriaceae in United States Children.

Direct observation of hole shift and characterization of spin states in radical ion pairs generated from photoinduced electron transfer of (phenothiazine)(n)-anthraquinone (n = 1, 3) dyads.

Cardiac mitochondria and reactive oxygen species generation.

Human granulocyte generation of hydroxyl radical.

Probing electronic communication for efficient light-harvesting functionality: dyads containing a common perylene and a porphyrin, chlorin, or bacteriochlorin.

A "click-chemistry" approach for the synthesis of porphyrin dyads as sensitizers for dye-sensitized solar cells.

Antenna effects in truxene-bridged BODIPY triarylzinc(II)porphyrin dyads: evidence for a dual Dexter-Förster mechanism.

Independent Generation and Reactivity of Thymidine Radical Cations.

Hydroxyl radical generation by red tide algae.

Ruthenium, osmium and rhodium complexes of 1,4-diaryl 1,4-diazabutadiene: radical versus non-radical states.

CNDO molecular orbital calculations on porphyrins--I. Ground and excited states of porphyrin, divinylporphyrin and tetraphenylporphyrin.

Independent generation and reactivity of uridin-2'-yl radical.

Generation of furazolidone radical anion and its inhibition by glutathione.

On the photo-induced charge-carrier generation within monolayers of self-assembled organic donor-acceptor dyads.

Generation and Detection of Reactive Oxygen Species in Photocatalysis.

On-chip generation of photon-triplet states.

CNDO molecular orbital calculations on porphyrins--II. Ground states of dilithium and disodium porphyrin.

Ascorbate both activates and inactivates bleomycin by free radical generation.

Characterization of Porphyrin-Co(III)-'Nitrene Radical' Species Relevant in Catalytic Nitrene Transfer Reactions.