Measuring Charge-Separation Dynamics via Oligomer Length

Article pubs.acs.org/JPCC

Measuring Charge-Separation Dynamics via Oligomer Length Variation Florian Kanal,† Stefan Ruetzel,† Han Lu,‡ Michael Moos,‡ Marco Holzapfel,‡ Tobias Brixner,*,†,§ and Christoph Lambert*,‡,§ †

Institut für Physikalische und Theoretische Chemie, ‡Institut für Organische Chemie, and §Center for Nanosystems Chemistry, Universität Würzburg, Am Hubland, 97074 Würzburg, Germany S Supporting Information *

ABSTRACT: We study the optically induced charge-transfer dynamics in donor−acceptor oligomers of different chain lengths. The combination of systematic synthesis, electrochemical measurements, and ultrafast transient absorption spectroscopy allows us to determine the charge-transfer properties and dynamics in donor−acceptor systems of confined lengths. Calculations within Marcus and Jortner electron-transfer theory explain the different charge-recombination times. For compounds in which complete charge separation can occur we deduce fast equilibration between different charge-transfer configurations prior to charge recombination. Thus, monoexponential charge-recombination kinetics are observed, as only the smallest-barrier configuration leads to relaxation to the ground state. The systematic oligomer length variation along with time-resolved spectroscopy allows us to determine how far apart charges can be separated in multichromophore donor−acceptor systems. Such information is relevant for understanding on a microscopic level the charge carrier mobility in materials for organic electronics and photovoltaics.

■

INTRODUCTION Optically induced charge transfer is at the heart of many dynamic processes in biology and technology such as in photosynthesis, photovoltaics, or optoelectronics in general. The efficiency of charge separation and propagation is often decisive for the overall function. Thus, charge separation is studied extensively both in natural1−8 and in artificial systems.9−16 In the microscopic domain, primary dynamics of charge transfer and recombination are accessible with femtosecond spectroscopy by monitoring the kinetic and spectral evolution of transient species.17−21 Additional information can be obtained from coherent two-dimensional spectroscopy,22−24 which recently has been applied to charge transfer.25,26 On the other hand, charge propagation in the macroscopic domain is studied by mobility measurements.27−30 In that case, one obtains information that represents an average over many microscopic dynamical steps. It is a significant and mostly unsolved challenge to predict carrier mobility and, thus, device efficiency from fundamental properties of the system’s constituent parts. Thus, it remains difficult to bridge the gap between the microscopic information obtained from femtosecond charge-transfer experiments and the macroscopically averaged observation of charge separation over larger distances. In particular, we would like to determine how far apart charges can move in (polymeric) molecular donor−acceptor systems, once the initial charge transfer from a donor to an (neighboring) acceptor has occurred. Do the separated charges © XXXX American Chemical Society

separate further, and what are the rate-determining steps for either recombination or charge transport? In the present work, we show how to address such questions by combining systematic synthesis and ultrafast spectroscopy. The key idea is to assemble donor−acceptor oligomers of increasing length up to a polymer and investigate their charge-recombination dynamics with transient absorption spectroscopy. Thus, if charges are indeed separated further and further along a donor−acceptor chain, the charge recombination measured for the increasingly longer chains should slow down because the probability for recombination drops with the distance between the charges. Comparison of time scales between the different systems of confined lengths will indicate if longer-range transport occurs or if recombination takes place immediately from the initially prepared charge-transfer state. Specifically, we are interested in low-band-gap polymers as they play an important role in the development of organic semiconductor devices such as organic bulk-heterojunction solar cells, organic field-effect transistors, and organic lightemitting devices.31−33 Impressive performances have been demonstrated in all these fields quite recently. For the design of low-band-gap polymers several approaches can be used.34 Rigidification of conjugation pathways, mixing of benzenoid Received: August 8, 2014 Revised: September 15, 2014

A

dx.doi.org/10.1021/jp508032k | J. Phys. Chem. C XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry C

Article

and quinoid structure in order to achieve bond-length equilibration, and the use of donor−acceptor copolymers are the main concepts to achieve this goal. Here we compare charge-separation and -recombination dynamics in donor (D)− acceptor (A) oligomers of systematically varied length and in a corresponding copolymer (P) as shown in Chart 1.

Constituents are the electron-accepting perchlorotriphenylmethyl radical (PCTM)35 and the electron-donating triarylamine (TAA).19,36−38 We used the PCTM moiety as this gives in combination with the triarylamine a donor−acceptor unit with a very low-lying band gap of about 1.0−1.2 eV. We apply electrochemistry, spectroelectrochemistry, femtosecond transient absorption spectroscopy, and Marcus-theory analysis to deduce the charge-separation dynamics.

■

Chart 1. Donor (D, green)−Acceptor (A, blue) Oligomers and the Polymer P Investigated in This Work

RESULTS Synthesis. For the synthesis (Supporting Information, Scheme S1) of model monomer DA and oligomers ADA and DAD we used the Horner−Wadsworth−Emmons reaction in THF with KOtBu as the base of appropriate aldehydes and phosphonates that have all previously been published.39,40 This leads to the DA-H, DAD-H, and ADA-H precursors, which were then characterized by NMR spectroscopic methods. In the case of DADA (Supporting Information, Scheme S2), the precursor of DADA was synthesized by Horner−Wadsworth− Emmons reaction from the two donor−acceptor fragments 5 and 6 which in turn were also prepared by Horner− Wadsworth−Emmons reaction from the appropriate aldehydes and phosphonates 1−4 in 1:1 stoichiometric ratio. Purification of all olefinic products 5, 6, ADA-H, DAD-H, and DADA-H was done by column chromatography and subsequent GPC. The NMR analytics of 5 and DADA-H proved to be complicated because both the triarylamines and the PCTM-H centers are chiral propellers that yield mixtures of diastereoisomers as products upon CC coupling. These diastereoisomers do not equilibrate at room temperature (RT) on the NMR time scale because of the relatively high barrier at the PCTM-H groups which should be even higher than that of a PCTM group (ΔG303 * = 98 kJ mol−1).41,42 The analysis is further hampered by an isotopic shift of the 13C signals caused by the 35Cl and 37Cl isotopes. Thus, in the experimental part (Supporting Information) we group and label signals that belong to chemically equivalent atoms. The DA-H, ADA-H, DAD-H, and DADA-H precursors were deprotonated at the PCTM center with n-butylammonium hydroxide and oxidized with p-chloranil to yield the radicals DA and DAD and the diradicals ADA and DADA. The purity of the radical products was checked by mass spectroscopy and by differential pulse voltammetry (DPV), which prove the appropriate ratio of oxidized TAA and reduced PCTM centers by integration of the signals. Electrochemistry. Electrochemical methods not only allow checking the purity of the radicals via integration of their reduction and oxidation waves in the differential pulse voltammograms (DPV, see Supporting Information, Figure S1) but also yield insight into interactions of charged systems. The redox potentials were measured by cyclic voltammetry (CV) in DCM/TBAPF6 solution (see Figure 1) and are given versus the ferrocene/ferrocenium redox couple in Table 1. The chemical reversibility of each process was ascertained by multisweep experiments under thin-layer conditions. Because of overlapping redox processes, the redox potentials were obtained by digital fitting of the cyclic voltammograms. Thereby, the two redox potentials of the reduction processes of ADA and of DADA and the two oxidation processes in DAD were determined. Only in DADA the two oxidation processes are separated enough to be clearly visible both in the cyclic voltammograms and in the differential pulse voltammetry. While it is tempting to interpret the different potential B



Article

UV/Vis/NIR-Absorption Spectroscopy. Absorption spectra of all radical oligomers were measured both in dichloromethane (DCM) and in toluene and are given in Figure 2. The spectra of all the radical oligomers are rather similar. While the band at about 35000 cm−1 is caused by localized transitions at the TAA,45 those between about 16000 and 30000 cm−1 are due to the PCTM.46 The most interesting band however is that at about 12000 cm−1, which is due to an optically induced charge transfer from the TAA donor to the PCTM acceptor and thus is called “intervalence charge-transfer” (IVCT) band.10,47 Because of the number of donor−acceptor interactions, this band must consist of one transition in DA, two transitions in ADA and DAD and of three in DADA. There is only little difference between spectra in DCM and toluene, the most remarkable difference being the somewhat broader IVCT band in DCM. The spectral data for the IVCT bands are collected in Table 2 together with the transition moments μeg calculated by integration of the IVCT band, μeg2 =

3hcε0 ln 10

9n 2000π NA (n + 2)2 2

∫ νε ̃ dν ̃

(1)

where h is Planck’s constant, c is the speed of light in vacuum (2.9979 × 108 m s−1), n is the refractive index of the solvent, NA is Avogadro’s constant, ε0 is the permittivity of the vacuum (8.8542 × 10−12 C2 J−1 m−1), and ε is the molar extinction coefficient (in units of M−1 cm−1). For this purpose, the IVCT band was modeled by three Gaussian functions in order to separate the band from the adjacent PCTM bands at higher energy. The square of the transition moments is proportional to the oscillator strength and the intensity of the transition. While we expected equal intensity for ADA and DAD there is about a 30% deviation. μ2eg of DADA is larger and that of DA smaller. From the number of donor−acceptor interactions one expects a 2 ratio of 1:2:2:3 for μeg of DA/ADA/DAD/DADA. The observed ratio is 1:1.5:2.2:2.6. The reason for these deviations from the theoretical values is presently unclear. UV/vis/NIR-spectroelectrochemistry measurements were carried out in DCM/TBAPF6 solution and are similar for all radicals. Thus, we present only those of the ADA radical in Figure 3. Upon oxidation new bands rise at 14200 cm−1, which is due to a localized transition within the TAA radical cation,45 and at 9000 cm−1, which is caused by an optically induced electron transfer from the neutral PCTM radical center to the TAA radical cation.36 Position and intensity of the bands are in agreement with those of the DA monomer. Upon reduction, a strong new band rises at about 18000 cm−1, which is caused by the PCTM anion.39,48 This band refers to two PCTM moieties as their redox potentials are too close to stop reduction at the monoanion of ADA. The intensity of this band is also in agreement with twice the intensity of the reduced DA monomer.19,36 In both cases, oxidation and reduction, the IVCT band at 12500 cm−1 disappears. These measurements will be helpful when we interpret the transient absorption spectra in the next section. Transient Absorption Spectroscopy. Excitation at a pump wavelength of 800 nm leads to direct population of the IVCT state in TAA−PCTM systems (“optical excitation”). Subsequent solvent reorganization leads to an increase of the excited-state absorption of the IVCT state. Depending on the solvent’s reorganization lifetimes and the charge recombination (CR) lifetimes in the DA-system under consideration, the combination of solvent relaxation and CR leads to the dynamics

Figure 1. Cyclic voltammograms of radicals (a) ADA, (b) DAD, and (c) DADA in DCM/TBAPF6 (0.2 M) vs Fc/Fc+. Scan rate v = 250 mV s−1.

Table 1. Redox Potentials of Radicals DA, ADA, DAD, and DADA; Redox Potentials of Terminal Groups are Given in Red and Those of Internal Groups in Bluea

a

Half-wave potentials vs Fc/Fc+ in DCM/TBAPF6 (0.2 M), scan rate v = 250 mV s−1.

separations of oxidative processes in DAD and reductive processes in ADA with differing electronic communications between the redox centers, we stress that ion pairing with the electrolyte counterions may strongly influence redox separations and, thus, these are no reliable estimates for interactions.43,44 In case of DADA, the individual redox centers in the two pairs of PCTM and TAA redox centers are not equivalent anyway, as one of each is terminal and the other within the molecule. Thus, the different neighborhood shifts the redox potential of the terminal TAA to lower values (+177 mV) and the terminal PCTM to higher (less negative) potentials (−676 mV). This is in agreement with the redox potentials of equivalent centers in the ADA and DAD systems. The redox potentials of the TAA and PCTM centers are in agreement with values found for polymer P.19 At potentials higher than about 800 mV, second oxidation processes of the TAA and possibly of the first oxidation of the PCTM occur (see Figure S1, Supporting Information, DPV). We refrain from giving exact potentials as these processes are chemically not fully reversible. C



Article

Figure 2. Absorption spectra of (a) ADA, (b) DAD, and (c) DADA in DCM and toluene at RT. (d) Absorption spectra of ADA, DAD, DADA, and DA in toluene.

selected difference spectra for various delay times τ after excitation in a spectral probe range from 13800 cm−1 (725 nm) to 24000 cm−1 (417 nm) for the molecules investigated in this study. The temporal evolution of the transient signals can be seen in Figure 5 for a set of transients in the region of groundstate bleaching (blue), absorption of PCTM− (green) and TAA+• (red). The evolution-associated difference spectra (EADS, Supporting Information, Figure S2) represent the amplitudes of the global-fitting-routine in a sequential model (Supporting Information). The lifetimes associated with the EADS are given in Table 3. In the cases where subsequent lifetimes are almost equal, the outcome of the fit was worse for one temporal component less, even if the relative error of these lifetimes gets larger than for the other lifetimes. Additionally, for the sake of consistency we used the same model for all investigated oligomers. For the monomer DA dissolved in DCM, one observes two distinct bands around 20000 cm−1 (500 nm) and below 15000 cm−1 (667 nm) after excitation (Figure 4a). These absorption bands of the IVCT state were assigned to excited-state transitions of the PCTM− and the TAA+• moiety, respectively. The assignment of the bands to IVCT-state transitions is supported by the spectroelectrochemistry data (Figure 3) where the absorption bands of an oxidized donor (TAA+•) and reduced acceptor (PCTM−) fit the observed transient absorption bands. The bands in the spectroelectrochemical data are slightly shifted toward lower wavenumbers [18000 cm−1 (555 nm) and 14200 cm−1 (704 nm)] compared to the time-resolved data. This red-shift was also observed for polymer P and methoxy-substituted DA.19 The shift might be explained by the influence of the solvent (DCM/TBAPF6) used for the spectroelectrochemistry measurements. The maximum of the TAA+• band cannot unambiguously be determined as for the red-edge part of the recorded spectrum, the bleaching of the PCTM band overlaps with the TAA+• absorption. The groundstate bleaching of DA is observed above 23500 cm−1. For longer delay times (≥1 ps), the PCTM− absorption band becomes more structured and develops into two slightly

Table 2. Spectral Data of the IVCT Band of the Radical Species in Toluene ν̃max/cm−1 ε/M−1 cm−1 μeg/D μ2eg/D2

DA

ADA

DAD

DADA

11200 4500 3.15 9.9

12200 6500 3.90 15.2

11600 8900 4.66 21.7

12000 10000 5.04 25.4

Figure 3. Spectroelectrochemistry of ADA in DCM/TBAPF6 solution at RT.

described below and discussed in the next section. The thermal CR in TAA−PCTM systems can be observed by measuring the absorption of the transiently formed TAA radical cation (TAA+•) and the PCTM anion (PCTM−) in the IVCT state. The absorption of TAA+• is reported to be between 14300 and 12500 cm−1 (700 and 800 nm),49 and the absorption of perchlorinated triarylcarbanions is around 19050 cm−1 (525 nm).50 The previously studied monomer DA with methoxysubstitution at the donor moiety and various related TAA− PCTM systems with different spacers, as well as polymer P show transient absorptions of the TAA+• and PCTM− species formed upon optical excitation into the IVCT state in the above-mentioned spectral regions.10,17−19,37,38 Figure 4 shows D



Article

Figure 4. Transient difference spectra for various pump−probe delay times. (a−d) Experimental data of DA, ADA, DAD, and DADA in DCM is shown in different colored lines for delay times τ > 0. (e−h) Experimental data of DA, ADA, DAD, and DADA in toluene is shown in different colored lines for delay times τ > 0.

tion upon photoexcitation.17 The decay of the transient absorption signals and the recovery of the ground-state bleaching were modeled with two exponentials resulting in τ2 = 450 fs and τ3 = 1.4 ps. The maximum of the TAA+• band shifts for delay times >2 ps to approximately 16000 cm−1 (625 nm). Even though the CR observed for the methoxysubstituted DA was determined to be monoexponential in our previous work,19 the two time constants for the decay of the IVCT state population in DA found in this analysis are consistent with the earlier results as those were not fitted globally but for selected wavelengths only. The difference spectra of ADA (Figure 4b) in DCM strongly resemble the transient absorption of DA. The bands of PCTM− and TAA+• as well as the ground-state bleaching are at the same

separated maxima. The center between both new maxima is around the spectral position of the maximum of the PCTM− band for a delay time of 500 fs (Figure 4a). Transients (Figure 5a) are shaded by the coherent artifact51,52 for delay times around time zero but show a finite rise time for both IVCTstate absorption bands (green and red circles) and the groundstate bleaching (blue circles). All signals decay and recover, respectively, within less than 10 ps. The lifetime of the rise of the transient IVCT absorption bands was determined by the global fitting routine to be 450 fs (τ1). This lifetime is in accordance with the rise time of the excited-state absorptions observed for the methoxy-substituted DA in DCM and methoxy-substituted DA without the bridging ethylene unit and corresponds to the previously studied solvent reorganizaE



Article

Figure 5. Transients for various probe wavelengths and corresponding global fits. (a−d) Experimental data of DA, ADA, DAD, and DADA in DCM is given in circles, the result of the global fit is depicted with lines. (e−h) Experimental data of DA, ADA, DAD, and DADA in toluene is given in circles, the result of the global fit is depicted with lines.

therefore on the same order of magnitude as in the monomer. The biexponential decay of the IVCT-state absorption bands was fitted with 910 and 930 fs. For the DAD oligomer the maximum of the PCTM− absorption band is centered at 18500 cm−1 (541 nm) due to a more pronounced overlap with a slightly red-shifted groundstate bleaching compared to ADA. The TAA+• absorption band remains maximal below 14500 cm−1 (690 nm). The blue-shift of the TAA+• band with temporal evolution is comparable to that in DA and ADA, whereas the lower-energy maximum of the split PCTM− band shows only a small shift of about 500 cm−1 for longer delay times (Figure 4c). The solvent reorganization dynamics leading to a rise in the IVCT excited-state absorptions was fitted with a lifetime of τ1 = 490 fs. However, the decay resembles more the DA dynamics than the ADA dynamics with lifetimes of 500 fs and 1.2 ps (Figure 5c). Compared to the other oligomers dissolved in DCM, DADA shows a more pronounced ground-state bleaching signal (Figure 4d) due to the slightly red-shifted ground-state absorption with respect to the other oligomers but is still above 23500 cm−1 (426 nm). The PCTM− band is initially centered at 19000 cm−1 (526 nm) and only a marginal blue-

Table 3. Lifetimes in DCM and Toluene from a Sequential Global Fita DCM

toluene

DA ADA DAD DADA P DA ADA DAD DADA P

τ1/ps

τ2/ps

τ3/ps

b

c

1.4 0.93e 1.2 1.5 1.5 11 48 20 21 27

0.45 0.29 0.49f 0.54h 0.19 0.18 0.16 0.18 0.25

0.45 0.91d 0.50g 0.56i 0.63j 2.8 4.7 4.0 3.5 8.4

τ4/ps

τ5/ps

69

∞

110

∞

a

The relative error of the lifetimes is below 3% if not stated otherwise. b Relative error, 13%. cRelative error, 13%. dRelative error, 18%. e Relative error, 17%. fRelative error, 25%. gRelative error, 25%. h Relative error, 14%. iRelative error, 15%. jRelative error, 5%

spectral position as in DA. The shift of the maxima observable for ADA is similar to the shift of the band of DA as a function of time. The lifetime of the rise of the transient IVCT absorption bands (Figure 5b) was fitted with τ1 = 290 fs and is F



Article

equivalent to the CR in DAD (τ3 = 21 ps) and the decay of the excited-state absorption is also monoexponential. For reasons of better comparison we performed a global lifetime analysis of the polymer data which was previously only fitted for single wavelengths,19 in addition to the oligomers (Table 3). In the previous study, the polymer was excited at 525 nm, for which reason a comparison of the oligomer data and the polymer data should only be done for delay times >1 ps, where possible excess energy in the polymer already has dissipated to the bath. For the dynamics in DCM four time components were fitted. Due to a coarse temporal step size and a higher energetic excitation in the measurements of P, the first lifetime τ1 could not unambiguously be fitted and did not improve the fitting result either, which is the reason why it was not considered in the fit. Therefore, only one short time constant τ2, assumed to represent the CR lifetime, was determined. The slow solvent reorganization component was again fitted with a lifetime τ3 = 1.5 ps. Two additional longer lifetimes were necessary to adequately fit the data. The lifetime τ4 = 69 ps represents the CR of previously completely separated charges (τ3 in ref 19). The lifetime τ5 set to infinity represents the permanent offset observed in the previous experiment. In toluene, the dynamics for P were fitted with five temporal components. τ1 and τ2 represent the solvent reorganization, τ3 the CR, τ4 the charge recombination from a prior completely charge-separated structure, and τ5 the remaining offset. The EADS of the global fitting results are given in Figure S3a,b (DCM and toluene, Supporting Information). This model requires an almost negligible exciton binding energy between positive and negative charges as otherwise charges would not dissociate along the polymer chain. A small exciton binding energy in P was indeed found by comparison of electrochemically and optically determined band gaps.19 However, the extent of charge separation could not be manifested in these experiments.

shift with time is observed after the second maximum gets visible on the blue edge of the PCTM− absorption band. Together with the fact that the PCTM− and the TAA+• absorption bands are rather flat, these observations, which were already made for P, indicate that DADA starts to show first properties already observed for the polymeric compound. The DCM solution of the dimer DADA (Figure 5d) possesses a solvent reorganization lifetime τ1 = 540 fs and a biexponential decay of the IVCT excited-state absorptions with τ2 = 560 fs and τ3 = 1.5 ps. In the following, the dynamics in toluene are discussed. In the monomer DA, two bands can be observed in the visible spectral regime around 20000 cm−1 (500 nm) and 14000 cm−1 (714 nm) in addition to a ground-state bleaching around 23500 cm−1 (426 nm, Figure 4e). Apparently, the less polar solvent toluene does not significantly change the spectral properties of the transient absorption for the PCTM− and the TAA+• band. Also the ground-state bleaching is only slightly shifted in accordance with the small shift in the steady-state absorption. As in DCM, the two transient bands are assigned to the absorption of the IVCT state. The coherent artifact again partially disguises the dynamics directly after time zero (Figure 5e). Anyhow, the dynamics of all investigated oligomers change significantly in toluene. The rise of the PCTM− and the TAA+• signals is slower than in DCM and was fitted biexponentially in toluene. In DA, the faster component is given with a lifetime of 190 fs. The slower component has a lifetime of 2.8 ps. Such a biexponential solvent reorganization process has been reported earlier for toluene. 53 Compared to DCM, the charge recombination is a factor of about 10 slower in toluene and also slower than the typical solvent relaxation observed in toluene. Therefore, the absorptions of the TAA+• and the PCTM− moiety decay monoexponentially (Figure 5e). In addition, no spectral shift in the bands’ maxima is observed (Figure 4e). The ground-state bleaching is centered around 23500 cm−1 (426 nm) in accordance with the low-energy shoulder of the linear absorption spectrum (Figure 2). The CR was fitted with a lifetime of 11 ps and is in accordance with earlier results for methoxy-substituted DA.18,19 In the case of ADA, the solvent relaxation lifetimes were determined to be 180 fs and 4.7 ps (Figure 5f). The PCTM− absorption around 20500 cm−1 (488 nm) does not show a spectral shift with time, neither does the TAA+• band above 17000 cm−1 (588 nm, Figure 4f). The PCTM− and TAA+• bands decay monoexponentially. The ground-state bleaching and the excited-state absorption signal contributions have not completely recovered and decayed, respectively, until a delay time greater than 100 ps. This qualitative finding is quantified in the slowest CR dynamics within all four molecules investigated with τ3 = 48 ps. For DAD the solvent relaxation lifetimes were determined to be 160 fs and 4.0 ps (Figure 5g). The CR is indeed a factor of 2.4 faster (τ3 = 20 ps) than in ADA. The PCTM− band appears significantly red-shifted around 18500 cm−1 (541 nm). Like in ADA, both the PCTM− absorption and the ground-state bleaching signal slightly shift to higher wavenumbers (Figure 4g). Finally, the dimer DADA shows solvent relaxation lifetimes of 180 fs and 3.5 ps (Figure 5h). The PCTM− absorption around 19000 cm−1 (526 nm) and the TAA+• absorption below 14500 cm−1 (690 nm) exhibit no spectral shift with time in toluene (Figure 4h). Like DA, DADA does not show a shift of the ground-state bleaching. The lifetime of the CR in DADA is

■

DISCUSSION In case of ADA and DADA, the ground and the IVCT state might adopt either singlet or triplet spin multiplicity. These states of different spin multipliticity are assumed to be almost degenerate in energy and not suppossed to show different absorption properties, neither in the ground nor in the excited state. Therefore, the following discussion will not consider any effect of the different spin multiplicities on the interpretation of the time-resolved data. After optically induced CT from the TAA donor to the PCTM acceptor, the surrounding solvent molecules have to rearrange to the new charge distribution in the IVCT state. In toluene solutions of all four molecules under investigation two time constants for the rise of the transient absorption bands of the IVCT state were determined with the global fitting routine. The signals’ evolutions can be described by two processes assigned to a fast librational motion and a slower diffusive contribution of the solvent molecules to fit the IVCT charge distribution.54 From time-resolved fluorescence measurements of coumarin C153 it is known that toluene shows a short component for the solvent reorganization with a lifetime of 370 fs and a longer component of 2.7 ps.53 As the biexponential risings of the transient absorption bands of all four oligomers show similar time constants (τ1 = 160−190 fs and τ2 = 2.8−4.7 ps), these time constants can clearly be assigned to the solvent relaxation after population of the IVCT state because the optically induced electron transfer takes place at the G



Article

Table 4. Radii of Donor and Acceptor Moieties, Donor−Acceptor Distances, Redox Potentials, Reorganization Energies, Free Energies, and Barriers for the DA Oligomers and Pa D+ADA− + −

rA/Å rD/Å dDA/Å Eox 1/2/V Ered 1/2/V λi/eV λo(toluene)/eV λ(toluene)/eV ΔG0IVCT(toluene)/eV ΔG*(toluene)/eV λo(DCM)/eV λ(DCM)/eV ΔG0IVCT(DCM)/eV ΔG*(DCM)/eV

+ −

+ −

···D+ADA−··· + −

−

+ −

+ −

DA

AD A

DA D

bent

stretched

D A DA

DA D A

DAD A

···D A ···

bent

stretched

5.62 4.81 12.42 0.24 −0.69 0.247 0.05 0.30 −1.30 0.85 0.62 0.87 −0.80 0.00

5.62 4.81 12.44 0.34 −0.58 0.175 0.05 0.23 −1.29 1.26 0.62 0.80 −0.80 0.00

5.62 4.81 12.43 0.20 −0.68 0.247 0.05 0.30 −1.25 0.76 0.62 0.87 −0.75 0.00

5.62 4.81 26.10 0.18 −0.68 0.247 0.07 0.32 −1.47 1.05 0.85 1.10 −0.79 0.02

5.62 4.81 30.91 0.18 −0.68 0.247 0.07 0.32 −1.51 1.12 0.88 1.13 −0.80 0.02

5.62 4.81 12.44 0.18 −0.74 0.247 0.05 0.30 −1.29 0.83 0.62 0.87 −0.79 0.00

5.62 4.81 12.44 0.30 −0.74 0.175 0.05 0.23 −1.41 1.57 0.62 0.80 −0.91 0.00

5.62 4.81 12.44 0.30 −0.68 0.175 0.05 0.23 −1.34 1.39 0.62 0.80 −0.85 0.00

5.62 4.81 12.44 0.34 −0.69 0.175 0.05 0.23 −1.40 1.54 0.62 0.80 −0.90 0.00

5.62 4.81 26.10 0.34 −0.69 0.175 0.07 0.25 −1.65 2.02 0.85 1.03 −0.97 0.00

5.62 4.81 30.91 0.34 −0.69 0.175 0.07 0.25 −1.68 2.11 0.88 1.06 −0.98 0.00

+

a

The radii rA and rD represent the radii of spheres with the same surface than the Connolly molecular surface (see text). Outer reorganization energies (λo) are calculated from eq 3, and inner reorganization energies are taken from literature (TAA) and calculated by DFT (PCTM, see Supporting Information).

equilibrium ground-state geometry. The detection of the fast electron transfer, directly populating the IVCT state, is in our case not only limited by the temporal resolution of the setup, but also by the solvent reorganization, which leads to a delayed evolution of the signals’ maxima when the solvent has relaxed along an averaged solvent reaction coordinate into the IVCTstate geometry. The CR, which was determined to occur in the Marcus inverted region for methoxy-substituted DA,17,19 is much slower than the initial electron transfer. The present oligomers were designed from subunits known to have small inner reorganization energies, anticipating high energy barriers for the electron back-transfer (slower CR) from the optically induced IVCT state. Two time constants were reported for the dielectric relaxation process of DCM.55 The faster one has a lifetime of about 400 fs in accordance with the result of τ1 = 290−540 fs for the rise of the transient absorptions of TAA+• and PCTM− in the oligomers in DCM. The slower component of 1.5 ps is equivalent or even slower than the third lifetime (τ3) determined with the global fit. Therefore, the CR is described by the second lifetime from the global fit ranging from 450− 910 fs (τ2) in DCM. With the CR being faster than the slow component of the solvent relaxation, the decays of the excitedstate absorptions show biexponential behavior. Using solvents of similar spectroscopic properties like DCM but different solvent relaxation times could substantiate this assumption. Unfortunately, this approach is restricted for the inspected oligomers due to their limited solubility in solvents other than DCM and toluene. Additionally, one has to take into account that the rise of the transient absorption bands is shaded for short delay times by the coherent artifact. As the CR is ultrafast in DCM, one could also assign the third lifetime to the cooling of a hot ground state within less than 1.5 ps. This assumption seems reasonable considering the shapes of EADS 3 (Figure S2a−d, Supporting Information), which show a reduced ground-state bleaching and a broad red-shifted absorption. Within transient absorption spectroscopy, we cannot unambiguously assign either solvent relaxation or absorption of a hot ground state to EADS 3.

As for DCM and toluene strongly differing CR dynamics could be observed, this fact was scrutinized in the framework of Marcus theory. For gaining a more detailed insight we need estimates for the free energies ΔG of the electron transfer processes and for the reorganization energies λ associated with the electron transfer where the latter consists of an inner (vibrational, λi) and an outer (solvent, λo) contribution. Therefore, we calculated the free energies ΔG0IVCT of the red IVCT states from the redox potentials Eox 1/2 and E1/2 (Table 1) 56 with the approach of Weller, 0 ΔG IVCT =

NAze ox NAe 2 red (E1/2 − E1/2 )− 1000 1000 × 4πε0 ⎡⎛ 1 1 ⎞⎛ 1 1⎞ 1 ⎤ ⎥ × ⎢⎜ + ⎟⎜ − ⎟ + ⎢⎣⎝ 2rD εs ⎠ εsdDA ⎥⎦ 2rA ⎠⎝ εr

(2)

with the elementary charge e and the number z of transferred electrons. The permittivities of the solvent in which the electrochemical measurements were performed and the optically induced electron transfer was studied in are given by εs and εr, respectively. The radii rD and rA of the TAA and PCTM moieties were estimated by radii of spheres with the same surface as the Connolly molecular surface calculated for the single TAA and PCTM subunits within ChemBio3D Ultra.57 The distance dDA between the donor and acceptor centers, assumed as point charges, were calculated with CAMB3LYP/6-31G* level of theory using Gaussian09.58 The inner and outer reorganization energies λi and λo for CR can in principle be extracted from absorption or emission spectra by fitting them with the model of Jortner.59−61 However, there are some limitations to this approach that complicate a band-shape analysis. Thus, calculated values for the inner reorganization energies λi were taken from the literature. Density functional theory (DFT) calculations revealed an inner reorganization energy of λi = 0.274 eV for a methoxy-substituted TAA and λi = 0.129 eV for a methylsubstituted TAA62 and were assumed for terminal and central TAA units, respectively. For the PCTM moiety, λi was calculated to be 0.22 eV.63 Combining these values results in H



Article

inner reorganization energies of [(0.129 + 0.22)/2] eV = 0.175 eV (for processes between the central TAA+• unit and PCTM−) and [(0.274 + 0.22)/2] eV = 0.247 eV (between a terminal TAA+• unit and PCTM−), respectively, for the CR in the oligomers of this study (Table 4). The outer reorganization energies λo were calculated from55,64−66 λo =

e2 ⎛ 1 1 ⎞⎛ 1 1 1 ⎞ + − ⎜ 2 − ⎟⎜ ⎟ εs ⎠⎝ 2rD 4πε0 ⎝ n 2rA dDA ⎠

(3)

where n is the refractive index of the solvent. The results for the free energy and the reorganization energy calculations are given in Table 4. For the dimer DADA, different possible IVCT configurations (e.g., D+ADA−, D+A−DA, etc.) were considered, but also two different conformers (stretched vs bent, Chart 2) for D+ADA−

Figure 6. Free energies of IVCT states and the ground state for a charge recombination in the optimal and inverted region within Marcus theory. The ground state is depicted by the black parabola. The blue parabola represents an IVCT state in the optimal region where the activation barrier ΔG* is zero. In the inverted region, the activation barrier depends on the energetic position of the IVCT state with respect to the ground state (red and green parabolas).

Chart 2. Schematic Representation of the Stretched and Bent Configuration of DADA

The CR lifetimes in DCM are about one to 2 orders of magnitude shorter than in toluene solutions, which indicates that in the latter solvent the activation barrier is significantly larger. The calculated free energies for the CR barrier ΔG* are shown in Table 4. By comparing the free energies and the reorganization energies one sees that −ΔG0IVCT > λ in toluene, that is, the back electron transfer is clearly in the Marcus inverted region (Figure 6, red and green IVCT parabolas), that is, the CR rate is slowed down although the exergonicity of the CR is increased compared to the normal region. Despite that the calculation of ΔG* is valuable in an estimation of the Marcus region, the size of the activation barrier does not fully explain the different CR dynamics observed experimentally. In the case of ADA, the CR lifetime increases by a factor of ∼4 compared to DA. This is obvious, as the activation barrier increases significantly for ADA (ΔG* = 1.26 eV) compared to DA (ΔG* = 0.85 eV). Within 0.1 eV, DAD (ΔG* = 0.76 eV) has the same barrier for the CR as DA. However, the CR is a factor of 2 slower for DAD (Table 3). A possible explanation for the discrepancy of τ3 in DAD could be the effect that the estimations of λo and ΔG are less reliable in toluene than in dipolar solvents. This is because dipolar solvents are better described by their permittivity εs than quadrupolar solvents like toluene, and thus, reorganization energies are estimated more reliably by eq 3 in DCM than in toluene. If dielectric constants67 are compared, toluene (ε = 2.379) appears to be significantly less polar than DCM (ε = 9.08), but if empirical ET(30) values for polarity68 are considered, toluene is described about as polar (ET(30) = 33.9 kcal/mol) as ethyl ether (ET(30) = 34.5 kcal/mol) and therefore shows about a similar value for polarity as DCM (ET(30) = 40.7 kcal/mol). For DADA, the CR lifetime in toluene is the same as in DAD, but the activation barriers for the different configurations vary from the same barrier size as DA to barrier sizes even larger than in ADA. The fact that in DADA only one CR lifetime is observed and not several different oneseach according to a barrier corresponding to a specific configurationshows that there must be fast prior equilibrations between the IVCT configurations (Scheme 1). These prior equilibria consist of either charge separation (ksep1−ksep3) to the

that differ in the terminal TAA−PCTM distance. The reorganization energy λ was previously determined for the methoxy-substituted DA by a Jortner fitting approach to be 0.59 eV in DCM36 which is in accordance with the estimated reorganization energy of 0.87 eV of DA in DCM, confirming the use of eq 3 for the calculations. Within Marcus theory, the free energy barrier ΔG* (Table 4) for the electron back-transfer was calculated via ΔG* =

0 (λ + ΔG IVCT )2 4λ

(4)

All calculations were also performed for the polymer P, that is, for the DADA configurations with the properties of central TAA or PCTM units only. Even though these calculated parameters are rather rough estimates and do not allow a quantitative analysis, the corresponding electron-transfer region in terms of Marcus theory can be predicted by comparing the free energies and the reorganization energies. If for DADA only configurations are taken into consideration where the electron and the hole are centered on adjacent TAA and PCTM subunits, that is, assuming that the CR predominantly occurs from these configurations (see below), the free energies approximately equal the reorganization energies (−ΔG0IVCT ≈ λ within 0.1 eV) for all oligomers in DCM. Hence, in DCM the electron back transfer is estimated to take place in the optimal region where the activation barrier ΔG* approaches zero (Figure 6, blue IVCT parabola). The optimal region effect is also confirmed by the calculated barriers ΔG* for the CR in DCM, which are zero for all oligomers and possible configurations. From these findings, one can explain why the CR lifetimes τ2 in DCM show almost no difference for all four molecules. Although the oligomers possess significant differences in their electron-donor and -acceptor strengths, a CR barrier ΔG* = 0 leads to the maximal CR rate (kCR = 1/τ2 = 1.1−2.2 × 1012 s−1) in all oligomers. I



Article ∞

Scheme 1. Schematic Representation of the Equilibration and Charge Splitting within DADAa

k CR = 4π 2hc 2V 2 ∑ j=0

exp( −S)S j j!

1 4πλokBT

⎡ (jhcν ̃ + λ + ΔG 0 )2 ⎤ v o IVCT ⎥ × exp⎢ − 4λokBT ⎦ ⎣

(6)

with the Huang−Rhys factor S = λ1/ν̃v, the quantum number j, and the averaged molecular mode wavenumber ν̃v. The longitudinal solvent relaxation time τL within the adiabtic parameter HA,69 HA =

a

Energy transfers via charge shifts are depicted with their rates kEN1−kEN3, charge separations with the rates ksep1−ksep3. Note that the equilibrium arrows only represent an averaged rate constant as the rates and the energy of the equilibrated species might differ depending on which DADA IVCT configuration was initially excited. The height of the excited species does not represent their relative energies.

8π 2hc 2V 2τL λo

takes into acccount the rate-limiting effect of the diffuse solvent motion on the CR, leading to a modified eq 6 for kCR,17 ∞

k CR = 4π 2hc 2V 2 ∑ j=0

exp(− S)S j j!

1 4πλokBT

⎡ (jhcν ̃ + λ + ΔG 0 )2 ⎤ v o IVCT ⎥ × exp⎢− 4λokBT ⎣ ⎦

completely charge-separated D+ADA− configuration or charge shifts (kEN1−kEN3). The charge shifts are equal to energy transfers, that is, exciton diffusion, because the IVCT excitation is transferred to a different donor−acceptor pair via electron and hole transport. These prior equilibria provoke that CR occurs only from the configuration with the smallest CR barrier, that is, with the shortest CR lifetime. Configuration D+A−DA has the smallest CR barrier, and hence, the fast prior equilibrations lead to CR only from this DADA configuration (k1, Scheme 1) and therefore to suppression of CR from the other configurations (k2−k4) as equilibration to the fastest CR channel is faster than the actual CR. Justification for the assumption about the relative speed of these processes can be obtained from calculating barriers for the charge shift. This barrier was calculated in ADA (A−D+A to AD+A−) and DAD (D+A−D to DA−D+) for electron transfer between degenerate “donor” and “acceptor” (Δ(ΔG0IVCT) = ΔG0 = 0, ΔG* = λ/4) to be 0.06 and 0.08 eV, respectively. These values are much smaller than the barriers for CR in all oligomers (|ΔG*| ≥ 0.76 eV). Thus, indeed charge diffusion is faster than CR. The equilibration between the different IVCT configurations cannot be detected experimentally as the transient absorptions of the equilibrating configurations are equal. It was observed already for P that the absorption of the completely charge-separated structure (···D+···A−···) does not differ from the configuration with neighboring electrons and holes (···D+A−···). To validate the calculated barriers, we calculated the corresponding rates kCR for a thermal electron back transfer and a given electronic coupling integral V. Typically, this can be done with the Marcus equation,55 k CR = 4π 2hc 2V 2

(7)

× [1 + HA(exp(− S)Sj /j! )]−1

(8)

Even though the Jortner model assumes diabatic free energy surfaces with zero electronic coupling, it yields correct results for calculations of adiabaticity systems with the electronic coupling being larger than zero.36,37,61 For our calculations, we used j = 0−20 and V = 2310 cm−1 (V = 2310 cm−1 in acetonitrile37). For the averaged vibrational mode, we assumed a typical value ν̃v = 1550 cm−1 ≙= 3.08 × 10−20 J = 0.19 eV (ν̃v = 1550 cm−1 for methoxy-DA in DCM36,37). The longitudinal solvent relaxation time (τL(DCM) = 400 fs, τL(toluene) = 370 fs) was taken from literature.53,55 The results of the rateconstant calculations are given in Table 5 as the corresponding CR lifetimes τCR = 1/kCR. Table 5. Calculated CR Lifetimes τCR in DCM and Toluene from Eq 8a DCM toluene

τCR/ps τCR/s

DA

ADA

DAD

DADA

P

0.26 6.6

0.26 8.7

0.27 3.0

0.26 6.1

0.25 5.8

a For DADA and P, only the configuration with the smallest barrier is considered.

For the optimal region in DCM the calculated lifetimes τCR (Table 5) match the experimentally determined CR lifetimes τ2 (Table 3) quite well, only deviating by a factor of 2. In toluene, the calculated lifetimes fit well for DA but differ for ADA, DAD, DADA, and P (τ3, Table 3). For ADA, DAD, DADA, and P, the lifetimes are significantly underestimated. This deviation is likely due to the quadrupolar nature of toluene, preventing accurate estimates of λo (eq 3) and thus of kCR (eq 8), as explained above. The dimer DADA is the smallest DA-based structure that allows complete charge separation along the molecular backbone (D+ADA−). With the equilibrations (Scheme 1) in the structure being significantly faster than the CR, one can explain why charge splitting in P leads to the observed additional time constant (Table 3), because the single equilibrations sum up to a rate-determining slow process. In the dimer DADA the charges cannot separate further apart, that is, no additional temporal component is observed. In polymer

⎡ (λ + λ + ΔG 0 )2 ⎤ 1 o IVCT ⎥ exp⎢ i 4πλokBT 4λokBT ⎦ ⎣ (5)

with the Boltzmann constant kB and the temperature T = 295.15 K. Especially in the inverted region (i.e., in toluene), the Marcus equation yields rate constants that underestimate the experimentally observed rate constants by orders of magnitude. Thus, the use of the semiclassical Jortner equation leads to more reliable rate constants J



Article

for optoelectronics. The combination of synthesis of variablelength DA oligomers with ultrafast spectroscopy, as demonstrated in the present work, allows determining whether photoinduced charge transfer occurs only to a neighboring subunit or the charges split further apart. The latter process is desirable for charge carrier mobility in photovoltaic devices, and thus, microscopic evaluations of charge transport are valuable tools.

P, the separated charges can even migrate further apart; therefore, a larger number of fast charge shifts sum up to the overall additional lifetime τ4. For P the calculated CR barriers are even larger than in ADA, but P exhibits a CR lifetime in between the CR lifetimes of DAD and ADA, which is again a hint that toluene is not well represented by its dielectric constant in eq 3.

■

■

CONCLUSION AND OUTLOOK We presented the synthesis, UV/vis/NIR linear absorption spectroscopy, electrochemistry, spectroelectrochemistry, and transient absorption spectroscopy of three different oligomers and a reference monomer based on TAA electron donors and PCTM electron acceptors. After optically induced electron transfer, these systems show charge recombination (CR) dynamics which we unraveled by ultrafast transient absorption spectroscopy. In combination with electron-transfer-theory calculations of the CR barriers ΔG*, and the CR lifetimes for the compounds, we could explain the experimentally observed CR lifetimes determined from a global fit of the data. The CR dynamics in DCM are superimposed by a slow component of the solvent reorganization, yielding a biexponential decay of the IVCT excited-state absorptions. In DCM, the CR is fast (∼500 fs) and does not show significant differences among the investigated oligomers. From this experimental finding and the calculated CR barriers being zero for all oligomers, we derive that the CR takes place in the optimal region, that is, without a barrier, the CR reaches the maximal possible rate. In toluene, however, the solvent relaxation lifetimes show up as a biexponential rise of the IVCT excited-state absorptions. The actual CR is much slower (11−48 ps) than the solvent reorganization. Hence, the CR is significantly slower than in DCM, taking place in the Marcus-inverted region, with a much larger CR barrier. The inverted-region effect is also confirmed by the calculated CR barriers being larger than the calculated reorganization energies. For the dimer DADA, only one CR lifetime was observed, and not several ones as one would expect from the different possible excited IVCT configurations. Therefore, we postulated fast prior equilibrations among the IVCT configurations. On the one hand, these equilibrations explain why only one CR lifetime is observed for DADA by the fast population of the lowest-barrier configuration outcompeting the slower CR channels. On the other hand, prior equilibrations make clear that already from a length of two DA pairs on, charge shifts and splittings exhibit only small barriers, that is, electron−hole pairs on linear TAA−PCTM compounds already separate in the smallest possible compound, allowing a completely chargeseparated structure. These fast, small-barrier processes then lead to the migration of charges in larger structures, the sum yielding the slower CR component observed in P. In the present study, we determined that charges already separate in DADA, the smallest molecular structure, where complete charge separation is possible. The access to defined longer DA oligomers is so far synthetically challenging but would give some additional information about how far mobile charges separate on a given molecular backbone. Powerdependent exciton−exciton annihilation experiments shown previously in molecular aggregates and polymer films70,71 would also give some more detailed insight. CR and charge splitting are important processes to study for the understanding of natural light harvesting-systems and low-band-gap materials

ASSOCIATED CONTENT

S Supporting Information *

Synthesis, characterization, experimental procedures, and additional data. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors acknowledge financial support from the German Research Foundation (DFG) within the Research Unit “LightInduced Dynamics in Molecular Aggregates” (FOR 1809) and La991/12-2 as well as from the European Cooperation in Science and Technology (COST) Action “PERSPECT-H2O” (CM1202). F.K. further acknowledges financial support from the “Fonds der Chemischen Industrie”.

■

REFERENCES

(1) Slavov, C.; Ballottari, M.; Morosinotto, T.; Bassi, R.; Holzwarth, A. R. Trap-Limited Charge Separation Kinetics in Higher Plant Photosystem I Complexes. Biophys. J. 2008, 94, 3601−3612. (2) Szczepaniak, M.; Sander, J.; Nowaczyk, M.; Müller, M. G.; Rögner, M.; Holzwarth, A. R. Charge Separation, Stabilization, and Protein Relaxation in Photosystem II Core Particles with Closed Reaction Center. Biophys. J. 2009, 96, 621−631. (3) Pawlowicz, N. P.; van Grondelle, R.; van Stokkum, I. H. M.; Breton, J.; Jones, M. R.; Groot, M. L. Identification of the First Steps in Charge Separation in Bacterial Photosynthetic Reaction Centers of Rhodobacter sphaeroides by Ultrafast Mid-Infrared Spectroscopy: Electron Transfer and Protein Dynamics. Biophys. J. 2008, 95, 1268−1284. (4) Zhu, J.; van Stokkum, I. H. M.; Paparelli, L.; Jones, M. R.; Groot, M. L. Early Bacteriopheophytin Reduction in Charge Separation in Reaction Centers of Rhodobacter sphaeroides. Biophys. J. 2013, 104, 2493−2502. (5) Di Donato, M.; Stahl, A. D.; van Stokkum, I. H. M.; van Grondelle, R.; Groot, M.-L. Cofactors Involved in Light-Driven Charge Separation in Photosystem I Identified by Subpicosecond Infrared Spectroscopy. Biochemistry 2011, 50, 480−490. (6) Romero, E.; van Stokkum, I. H. M.; Novoderezhkin, V. I.; Dekker, J. P.; van Grondelle, R. Two Different Charge Separation Pathways in Photosystem II. Biochemistry 2010, 49, 4300−4307. (7) Espagne, A.; Changenet-Barret, P.; Baudin, J.-B.; Plaza, P.; Martin, M. M. Photoinduced Charge Shift as the Driving Force for the Excited-State Relaxation of Analogues of the Photoactive Yellow Protein Chromophore in Solution. J. Photochem. Photobiol., A 2007, 185, 245−252. (8) Zinth, W.; Wachtveitl, J. The First Picoseconds in Bacterial PhotosynthesisUltrafast Electron Transfer for the Efficient Conversion of Light Energy. ChemPhysChem 2005, 6, 871−880.

K



Article

(9) Petersson, J.; Eklund, M.; Davidsson, J.; Hammarström, L. Ultrafast Electron Transfer Dynamics of a Zn(II) Porphyrin−Viologen Complex Revisited: S2 vs S1 Reactions and Survival of Excess Excitation Energy. J. Phys. Chem. B 2010, 114, 14329−14338. (10) Heckmann, A.; Lambert, C. Organic Mixed-Valence Compounds: A Playground for Electrons and Holes. Angew. Chem., Int. Ed. 2012, 51, 326−392. (11) Akemann, W.; Laage, D.; Plaza, P.; Martin, M. M.; BlanchardDesce, M. Photoinduced Intramolecular Charge Transfer in Push−Pull Polyenes: Effects of Solvation, Electron-Donor Group, and Polyenic Chain Length. J. Phys. Chem. B 2008, 112, 358−368. (12) Bhosale, S.; Sisson, A. L.; Talukdar, P.; Fürstenberg, A.; Banerji, N.; Vauthey, E.; Bollot, G.; Mareda, J.; Röger, C.; Würthner, F.; et al. Photoproduction of Proton Gradients with π-Stacked Fluorophore Scaffolds in Lipid Bilayers. Science 2006, 313, 84−86. (13) Wasielewski, M. R. Self-Assembly Strategies for Integrating Light Harvesting and Charge Separation in Artificial Photosynthetic Systems. Acc. Chem. Res. 2009, 42, 1910−1921. (14) Villamaina, D.; Kelson, M. M. A.; Bhosale, S. V.; Vauthey, E. Excitation Wavelength Dependence of the Charge Separation Pathways in Tetraporphyrin-Naphthalene Diimide Pentads. Phys. Chem. Chem. Phys. 2014, 16, 5188−5200. (15) Letrun, R.; Vauthey, E. Excitation Wavelength Dependence of the Dynamics of Bimolecular Photoinduced Electron Transfer Reactions. J. Phys. Chem. Lett. 2014, 5, 1685−1690. (16) Rosspeintner, A.; Angulo, G.; Vauthey, E. Bimolecular Photoinduced Electron Transfer Beyond the Diffusion Limit: The Rehm−Weller Experiment Revisited with Femtosecond Time Resolution. J. Am. Chem. Soc. 2014, 136, 2026−2032. (17) Maksimenka, R.; Margraf, M.; Köhler, J.; Heckmann, A.; Lambert, C.; Fischer, I. Femtosecond Dynamics of Electron Transfer in a Neutral Organic Mixed-Valence Compound. Chem. Phys. 2008, 347, 436−445. (18) Dümmler, S.; Roth, W.; Fischer, I.; Heckmann, A.; Lambert, C. Excited-State Dynamics in a Neutral Organic Mixed-Valence Compound. Chem. Phys. Lett. 2005, 408, 264−268. (19) Reitzenstein, D.; Quast, T.; Kanal, F.; Kullmann, M.; Ruetzel, S.; Hammer, M. S.; Deibel, C.; Dyakonov, V.; Brixner, T.; Lambert, C. Synthesis and Electron Transfer Characteristics of a Neutral, LowBand-Gap, Mixed-Valence Polyradical. Chem. Mater. 2010, 22, 6641− 6655. (20) Mohammed, O. F. Ultrafast Intramolecular Charge Transfer of Formyl Perylene Observed Using Femtosecond Transient Absorption Spectroscopy. J. Phys. Chem. A 2010, 114, 11576−11582. (21) Kovalenko, S. A.; Eilers-König, N.; Senyushkina, T. A.; Ernsting, N. P. Charge Transfer and Solvation of Betaine-30 in Polar Solvents A Femtosecond Broadband Transient Absorption Study. J. Phys. Chem. A 2001, 105, 4834−4843. (22) Hybl, J. D.; Albrecht, A. W.; Gallagher Faeder, S. M.; Jonas, D. M. Two-Dimensional Electronic Spectroscopy. Chem. Phys. Lett. 1998, 297, 307−313. (23) Brixner, T.; Stenger, J.; Vaswani, H. M.; Cho, M.; Blankenship, R. E.; Fleming, G. R. Two-Dimensional Spectroscopy of Electronic Couplings in Photosynthesis. Nature 2005, 434, 625−628. (24) Cho, M. Coherent Two-Dimensional Optical Spectroscopy. Chem. Rev. 2008, 108, 1331−1418. (25) Bixner, O.; Lukeš, V.; Mančal, T.; Hauer, J.; Milota, F.; Fischer, M.; Pugliesi, I.; Bradler, M.; Schmid, W.; Riedle, E.; et al. Ultrafast Photo-Induced Charge Transfer Unveiled by Two-Dimensional Electronic Spectroscopy. J. Chem. Phys. 2012, 136, 204503− 204503−12. (26) Consani, C.; Auböck, G.; Mourik, F. v.; Chergui, M. Ultrafast Tryptophan-to-Heme Electron Transfer in Myoglobins Revealed by UV 2D Spectroscopy. Science 2013, 339, 1586−1589. (27) Kepler, R. G. Charge Carrier Production and Mobility in Anthracene Crystals. Phys. Rev. 1960, 119, 1226−1229. (28) Gailberger, M.; Bässler, H. DC and Transient Photoconductivity of Poly(2-phenyl-1,4-phenylenevinylene). Phys. Rev. B 1991, 44, 8643−8651.

(29) Meyer, H.; Haarer, D.; Naarmann, H.; Hörhold, H. H. Trap Distribution for Charge Carriers in Poly(paraphenylene vinylene) (PPV) and its Substituted Derivative DPOP-PPV. Phys. Rev. B 1995, 52, 2587−2598. (30) Veres, J.; Ogier, S.; Lloyd, G.; de Leeuw, D. Gate Insulators in Organic Field-Effect Transistors. Chem. Mater. 2004, 16, 4543−4555. (31) Bundgaard, E.; Krebs, F. C. Low Band Gap Polymers for Organic Photovoltaics. Sol. Energy Mater. Sol. Cells 2007, 91, 954−985. (32) Li, Y.; Pan, Z.; Miao, L.; Xing, Y.; Li, C.; Chen, Y. FluoroBenzoselenadiazole-Based Low Band Gap Polymers for High Efficiency Organic Solar Cells. Polym. Chem. 2013, 5, 330−334. (33) Cui, W.; Wudl, F. Dithienylbenzodipyrrolidone: New Acceptor for Donor−Acceptor Low Band Gap Polymers. Macromolecules 2013, 46, 7232−7238. (34) Cheng, Y.-J.; Yang, S.-H.; Hsu, C.-S. Synthesis of Conjugated Polymers for Organic Solar Cell Applications. Chem. Rev. 2009, 109, 5868−5923. (35) Ratera, I.; Veciana, J. Playing with Organic Radicals as Building Blocks for Functional Molecular Materials. Chem. Soc. Rev. 2011, 41, 303−349. (36) Heckmann, A.; Lambert, C.; Goebel, M.; Wortmann, R. Synthesis and Photophysics of a Neutral Organic Mixed-Valence Compound. Angew. Chem., Int. Ed. 2004, 43, 5851−5856. (37) Heckmann, A.; Lambert, C. Neutral Organic Mixed-Valence Compounds: Synthesis and All-Optical Evaluation of ElectronTransfer Parameters. J. Am. Chem. Soc. 2007, 129, 5515−5527. (38) Heckmann, A.; Dümmler, S.; Pauli, J.; Margraf, M.; Köhler, J.; Stich, D.; Lambert, C.; Fischer, I.; Resch-Genger, U. Highly Fluorescent Open-Shell NIR Dyes: The Time-Dependence of Back Electron Transfer in Triarylamine-Perchlorotriphenylmethyl Radicals. J. Phys. Chem. C 2009, 113, 20958−20966. (39) Rovira, C.; Ruiz-Molina, D.; Elsner, O.; Vidal-Gancedo, J.; Bonvoisin, J.; Launay, J.-P.; Veciana, J. Influence of Topology on the Long-Range Electron-Transfer Phenomenon. Chem.Eur. J. 2001, 7, 240−250. (40) Hua, J.; Meng, F.; Li, J.; Ding, F.; Fan, X.; Tian, H. Synthesis and Characterization of New Highly Soluble and Thermal-Stable Perylene−PPV Copolymers Containing Triphenylamine Moiety. Eur. Polym. J. 2006, 42, 2686−2694. (41) Veciana, J.; Ratera, I. In Stable Radicals: Fundamentals and Applied Aspects of Odd-Electron Compounds; Hicks, R., Ed.; Wiley: Oxford, 2010; pp 33−80. (42) Veciana, J.; Rovira, C.; Crespo, M. I.; Armet, O.; Domingo, V. M.; Palacio, F. Stable Polyradicals with High-Spin Ground States. 1. Synthesis, Separation, and Magnetic Characterization of the Stereoisomers of 2,4,5,6-Tetrachloro-α,α,α′,α′-Tetrakis(pentachlorophenyl)m-xylylene biradical. J. Am. Chem. Soc. 1991, 113, 2552−2561. (43) Barrière, F.; Geiger, W. E. Use of Weakly Coordinating Anions to Develop an Integrated Approach to the Tuning of ΔE1/2 Values by Medium Effects. J. Am. Chem. Soc. 2006, 128, 3980−3989. (44) Steeger, M.; Lambert, C. Charge-Transfer Interactions in TrisDonor−Tris-Acceptor Hexaarylbenzene Redox Chromophores. Chem.Eur. J. 2012, 18, 11937−11948. (45) Amthor, S.; Noller, B.; Lambert, C. UV/Vis/NIR Spectral Properties of Triarylamines and Their Corresponding Radical Cations. Chem. Phys. 2005, 316, 141−152. (46) Ballester, M. In Adv. Phys. Org. Chem.; Bethell, D., Ed.; Academic Press: London, 1989; Vol. 25; pp 267−445. (47) Hankache, J.; Wenger, O. S. Organic Mixed Valence. Chem. Rev. 2011, 111, 5138−5178. (48) Sporer, C.; Ratera, I.; Ruiz-Molina, D.; Zhao, Y.; Vidal-Gancedo, J.; Wurst, K.; Jaitner, P.; Clays, K.; Persoons, A.; Rovira, C.; et al. A Molecular Multiproperty Switching Array Based on the Redox Behavior of a Ferrocenyl Polychlorotriphenyhnethyl Radical. Angew. Chem., Int. Ed. 2004, 43, 5266−5268. (49) Schmidt, W.; Steckhan, E. Ü ber organische Elektronenüberträgersysteme, I. Elektrochemische und spektroskopische Untersuchung bromsubstituierter Triarylamin-Redoxsysteme. Chem. Ber. 1980, 113, 577−585. L



Article

(50) Ballester, M.; Pascual, I. Spin-Charge Exchange in a Stable Radical-Carbanion, and Related Intermolecular One-Electron Transfers. Tetrahedron Lett. 1985, 26, 5589−5592. (51) Ekvall, K.; van der Meulen, P.; Dhollande, C.; Berg, L.-E.; Pommeret, S.; Naskrecki, R.; Mialocq, J.-C. Cross Phase Modulation Artifact in Liquid Phase Transient Absorption Spectroscopy. J. Appl. Phys. 2000, 87, 2340. (52) Lorenc, M.; Ziolek, M.; Naskrecki, R.; Karolczak, J.; Kubicki, J.; Maciejewski, A. Artifacts in Femtosecond Transient Absorption Spectroscopy. Appl. Phys. B: Lasers Opt. 2002, 74, 19−27. (53) Reynolds, L.; Gardecki, J. A.; Frankland, S. J. V.; Horng, M. L.; Maroncelli, M. Dipole Solvation in Nondipolar Solvents: Experimental Studies of Reorganization Energies and Solvation Dynamics. J. Phys. Chem. 1996, 100, 10337−10354. (54) Carter, E. A.; Hynes, J. T. Solvation Dynamics for an Ion Pair in a Polar Solvent: Time-Dependent Fluorescence and Photochemical Charge Transfer. Chem. Phys. 1991, 94, 5961−5979. (55) Heitele, H. Dynamic Solvent Effects on Electron-Transfer Reactions. Angew. Chem., Int. Ed. 1993, 32, 359−377. (56) Weller, A. Photoinduced Electron Transfer in Solution: Exciplex and Radical Ion Pair Formation Free Enthalpies and their Solvent Dependence. Z. Phys. Chem. (Muenchen, Ger.) 1982, 133, 93−98. (57) ChemBio3D, Ultra 3.0; Perkin-Elmer: Waltham, MA, 2010. (58) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A. et al. Gaussian 09, Revision B.01; Gaussian Inc.: Wallingford, CT, 2009. (59) Gould, I. R.; Noukakis, D.; Gomez-Jahn, L.; Young, R. H.; Goodman, J. L.; Farid, S. Radiative and Nonradiative Electron Transfer in Contact Radical-Ion Pairs. Chem. Phys. 1993, 176, 439−456. (60) Cortes, J.; Heitele, H.; Jortner, J. Band-Shape Analysis of the Charge-Transfer Fluorescence in Barrelene-Based Electron Donor− Acceptor Compounds. J. Phys. Chem. 1994, 98, 2527−2536. (61) Nelsen, S. F.; Ramm, M. T.; Wolff, J. J.; Powell, D. R. Intramolecular Electron Transfer between Doubly Four σ-BondLinked Tetraalkylhydrazine Cationic and Neutral Units. J. Am. Chem. Soc. 1997, 119, 6863−6872. (62) Cias, P.; Slugovc, C.; Gescheidt, G. Hole Transport in Triphenylamine Based OLED Devices: From Theoretical Modeling to Properties Prediction. J. Phys. Chem. A 2011, 115, 14519−14525. (63) The inner reorganization energy of the PCTM moiety was calculated by DFT, B3LYP, 6-31G* using Gaussian09, analogous to ref 62. (64) Marcus, R. A. On the Theory of Oxidation-Reduction Reactions Involving Electron Transfer. I. J. Chem. Phys. 1956, 24, 966−978. (65) Truong, Thu Ba Charge Transfer to a Solvent State. 9. Effect of Solvent Reorganization Energy on Electron-Transfer Reaction Rate Constant. The Marcus Inverted Region. J. Phys. Chem. 1984, 88, 3906−3913. (66) Brunschwig, B. S.; Ehrenson, S.; Sutin, N. Solvent Reorganization in Optical and Thermal Electron-Transfer Processes. J. Phys. Chem. 1986, 90, 3657−3668. (67) Weast, R. C.; Lide, D. R. CRC Handbook of Chemistry and Physics, 70th ed.; CRC Press: Boca Raton, FL, 1989. (68) Reichardt, C. Solvatochromic Dyes as Solvent Polarity Indicators. Chem. Rev. 1994, 94, 2319−2358. (69) Jortner, J.; Bixon, M. Intramolecular Vibrational Excitations Accompanying Solvent-Controlled Electron Transfer Reactions. J. Chem. Phys. 1988, 88, 167. (70) Marciniak, H.; Li, X.-Q.; Würthner, F.; Lochbrunner, S. OneDimensional Exciton Diffusion in Perylene Bisimide Aggregates. J. Phys. Chem. A 2011, 115, 648−654. (71) Zaushitsyn, Y.; Jespersen, K. G.; Valkunas, L.; Sundström, V.; Yartsev, A. Ultrafast Dynamics of Singlet-Singlet and Singlet-Triplet Exciton Annihilation in Poly(3,2′-methoxy-5′-octylphenyl)thiophene Films. Phys. Rev. B 2007, 75, 195201.

M


Measuring Charge-Separation Dynamics via Oligomer Length

Recommend Documents