Acceptor Charge Transfer Yield - The

Article pubs.acs.org/JPCC

Hot Excitons Increase the Donor/Acceptor Charge Transfer Yield Michael Schulze,*,†,‡ Marc Han̈ sel,† and Petra Tegeder*,† †

Physikalisch-Chemisches Institut, Ruprecht-Karls-Universität Heidelberg, Im Neuenheimer Feld 253, 69120 Heidelberg, Germany Fachbereich Physik, Freie Universität Berlin, Arnimallee 14, 14195 Berlin, Germany

‡

S Supporting Information *

ABSTRACT: Understanding the photoinduced ultrafast charge transfer (CT) dynamics across the donor/acceptor interface is a prerequisite for optimizing the performance of organic photovoltaic devices. Time-resolved second harmonic generation, an interface-sensitive probe with femtosecond temporal resolution, is applied to investigate the well-defined single heterojunction C60/P3HT. The de-excitation of hot singlet excitons in the conduction bands of the polymer into localized excitonic states is observed. In the presence of the electron acceptor, the ultrafast population of a CT state is identified as the dominating relaxation channel. Interestingly, the charge transfer yield correlates with the excitation wavelength and rises with increasing excess energy.

1. INTRODUCTION Organic photovoltaic cells (OPVs) based on semiconducting polymers or molecules offer potential advantages over inorganic modules, such as lightweight, low-cost fabrication and flexibility.1 Their realization using a single active layer, however, suffers from low external quantum efficiencies2 due to small excitonic diffusion coefficients in contrast to significantly larger film thicknesses necessary for efficient optical absorption.3 The incorporation of fullerene derivatives as electron acceptors into the active layer is considered to circumvent this problem2 by enabling to tailor the donor and acceptor domain sizes via the preparation parameters and by drastically increasing the area of the interface.4 Since the first reported realization of a solar cell based on a polymer:fullerene blend in 1995,5 numerous studies have been published on photovoltaic devices based on the bulkheterojunction concept. One of the most prominent donor/ acceptor (D/A) systems is P3HT:PCBM (poly(3-hexylthiophene):phenyl-C61-butyric acid methyl ester).6 However, the widespread reported efficiencies of devices based on this prototypal blend,1 all being well below the values reached by state-of-the-art inorganic multijunction solar cells,7 indicate a demand on further investigations of the underlying ultrafast photoinduced dynamics at the D/A interface. In particular, the role of excess energy from above-bandgap excitation currently experiences revision from being generally considered to be lost in waste heat.8 On the one hand, recent findings indicate the crucial role of hot excitons: (1) for populating the charge transfer (CT) states,9 (2) to overcome the poorly screened Coulomb barrier in organic semiconductors and avoid trapping in relaxed CT ground states localized at the interface,10 and (3) to enhance the probability of charge dissociation out of delocalized CT states.11 Furthermore, although theoretical modeling is generally difficult due to the large system sizes,10 © 2014 American Chemical Society

wave function-based ab initio calculations recently revealed that the experimentally observed delocalization of excitonic states accompanying the high energetic excitation9 is crucial for the stabilization of the CT states.12 On the other hand, for the ultrafast charge separation into band-like acceptor states in blends of small molecules and polymers with fullerenes, respectively, no indication for the need of thermal (i.e., in the sense of Onsager-like models13) excess energy has been found.14 Furthermore, no correlation between the excitation energy and the external quantum efficiency of a variety of OPV model systems has been found in recent extensive studies.15,16 In the present work, we investigate the charge transfer across the donor/acceptor interface, which constitutes a central process in the light-to-current conversion in OPVs. By applying time-resolved second harmonic generation (TR-SHG) this process is investigated with a temporal resolution below 50 fs and excitation energy resolved. The CT from the initially excited donor P3HT to the electron acceptor C60 is found to occur on an ultrafast time scale. Most interestingly, the CT yield correlates with the excitation wavelength and rises with increasing excess energy.

2. EXPERIMENTAL SECTION As illustrated in Figure 1, TR-SHG probes the second order nonlinear optical (NLO) properties of a medium described by the susceptibility tensor χ(2) (or its molecular equivalent the first hyperpolarizability β).17 The quadratic dependency of the SHG signal on the strong electric field of the probing femtosecond laser pulse results in the intrinsic interface and Received: October 24, 2014 Revised: November 8, 2014 Published: November 14, 2014 28527

dx.doi.org/10.1021/jp510701w | J. Phys. Chem. C 2014, 118, 28527−28534

The Journal of Physical Chemistry C

Article

vacuum evaporation of the not-soluble C60 instead of spincoating of PCBM to reduce diffusion of acceptor molecules into the donor layer. The success of this preparation step is justified via polarization resolved TR-SHG measurements sensitive to interfacial roughness and inhomogeneity (see Supporting Information, Figure S1). Second, it has been demonstrated that neat P3HT of high regioregularity as here investigated exhibits the characteristic vibronic pattern in the absorption profile associated with the π-stacking-mediated self-organized lamella superstructure26 and accordingly inheres high transport capabilities even without prior annealing25 in contrast to the blends. In summary, the here investigated well-defined D/A interface between aggregated P3HT and vacuum-deposited C60 corresponds to the domain boundaries between the crystallized donor polymer and clustered fullerene regions, the latter wellseparated via an additional annealing step in the blends. For the SHG probe beam a wavelength of 800 nm has been chosen. The absorbance of both investigated compounds is neglectable at the corresponding photon energy of 1.55 eV.27,28 The pump wavelength was varied from 575 to 652 nm, which corresponds to photon energies between 2.15 and 1.90 eV. Although the latter are still below the optical band gap energy of the acceptor,29 the absorbance of C60 is approximately 1 order of magnitude larger at these wavelengths in comparison to the probe beam wavelength,28 i.e., sufficiently distinct to enable pumpprobe SHG measurements as discussed below. In P3HT these wavelengths are suitable to excite the S0 → S1 and S0 → S2 transitions (vide infra). Pump beam powers were kept below 30 μJ/cm2 close to the solar exposure20 and thus sufficiently low to avoid beam damage and to exclude undesired effects as excitonexciton annihilation3,9,30−32 or even thermally induced absorbance shifts.33

Figure 1. Schematic illustration of the time-resolved second harmonic generation (TR-SHG) experiment. A pump pulse generates hot, delocalized excitons in the semiconducting polymer P3HT (bottom). It thus induces a distortion in the interfacial electron density to which the second order nonlinear susceptibility χ(2) probed in SHG measurements is correlated. The subsequent formation of interfacial electric fields due to charge transfer to the electron acceptor C60 (top) results in electric field enhanced second harmonic generation (EFISH). Varying the pumpprobe delay Δt allows to monitor the temporal evolution of the exciton population.

surface sensitivity of the method: for homogeneous samples only the symmetry breaking interfaces contribute to the SHG signal.18 As the χ(2)-tensor reflects the electronic structure of the system, the SHG probe is sensitive to the dynamic distortion of the interfacial electron density, induced by the excitation of electronhole pairs by a pump laser beam. In the D/A heterojunction the energetic relaxation is in competition with the charge transfer across the interface, which is accompanied by the formation of an electric field between the adjacent materials. The according amplification of the TRSHG signal is known as electric field enhanced second harmonic generation (EFISH).19 Samples were prepared by vacuum evaporation of C60 (99% purity, MER, Figure 1 top) and spin coating of regioregular P3HT (regioregularity > 98%, average molecular weight < 50,000 MW, BASF Sepiolid P200, Figure 1 bottom) onto optically polished sapphire substrates (CrysTec, (0001)orientation). Pristine C60 and P3HT samples with thicknesses of 20 and 50 nm, respectively, were used as references for the C60/P3HT heterojunction. The latter was prepared by evaporating 12 nm C60 on top of a 50 nm thick P3HT absorption layer. Stacked systems were selected instead of bulk-heterojunctions to reduce the complexity of the investigated system to a single, well-defined interface. This facilitates an unambiguous interpretation of the measured TR-SHG data, as in bulkheterojunctions a recombination associated with interrupted percolation pathways and cross-currents of electrons and holes cannot be excluded.20 However, the photoinduced dynamics at the D/A interface of this model system can be expected to be very similar to those in P3HT:PCBM blends, as outlined in the following. It has to be considered that, in order to obtain wellperforming OPVs based on the bulk-heterojunction concept, an annealing step is crucial to achieve:21 (1) a phase-separation that results in the formation of domains of donor and acceptor molecules22,23 with a large interfacial area,4 (2) the crystallization of the polymer phase and hence improved transport properties,24,25 and (3) low ohmic contacts between the active layer and the electrodes and accordingly a reduced series resistance.21 Regarding the comparability of the photoinduced dynamics in our model system and in the bulk-heterojunctions, the first two points are of crucial importance. First, we chose

3. RESULTS AND DISCUSSION Figure 2 shows the TR-SHG data of pristine P3HT. The pumpinduced excitation at time zero leads to a decrease in the SHG

Figure 2. TR-SHG data of pristine P3HT. At time zero, the pump beam induces a pronounced decrease in the NLO response, which returns to the initial level within a few picoseconds. The fit (green) is the iterative convolution of the cross-correlation with a biexponential function, yielding τ1 = 154 ± 6 fs and τ2 = 1370 ± 80 fs. The changes in the NLO response are attributed to the excitation and subsequent thermalization and localization of hot singlet excitons.

signal amplitude, which returns to the initial level within a few picoseconds. The data are fitted (green curve) by iteratively convoluting the Gaussian profile of the pumpprobe crosscorrelation with a biexponential function yielding τ1 = 154 ± 6 fs and τ2 = 1370 ± 80 fs. Numerous studies have been conducted for the polymer, mostly based on time-resolved 28528



Article

π-stacking-mediated self-organized lamella superstructure of regioregular (RR) P3HT.4 Obviously, 652 nm is the only applied wavelength insufficient for the population of the S2 band. Accordingly, the fast relaxation channel is associated with the hot exciton thermalization in the energetically lower S1 band. The slow decay channel needs to be associated with the S2 band. Nevertheless, several possibilities for the detailed underlying relaxation mechanism are conceivable. First of all, Brown et al. proposed that the S0 → S1 transition originates from an intrachain excitation, while the S0 → S2 excitation is associated with an interchain exciton.44 In this regard, two differently (de)localized species of electronhole pairs would be excited and the localization within one and among adjacent chains would be the origin of the two observed time-constants, respectively. However, other groups report on a significantly improved agreement between experimentally observed and theoretically modeled absorption spectra by assuming an interchain excitation in the two-dimensional lamella aggregate structure regardless of the excitation energy, i.e., the existence of only one excitonic species.45−47 The slow relaxation channel accordingly needs to be associated with thermalization of excess vibronic energy (resulting from the FranckCondon transition45) within the S2 band and via the subsequent S2 → S1 transition. For pristine C60 (Figure 4) the TR-SHG signal amplitude rises during the excitation indicated by the cross-correlation

transient absorption spectroscopy. Consensus on the nature of the initial excitation apparently exists, that singlet excitons are immediately excited,20,34−38 although in some reports the possibility of an additional instantaneous population of polarons,23,37 polaron pairs,34,38 and even free charges35 is discussed. Regarding the subsequent dynamics, the formation of interchain polarons not immediately but within less than one picosecond out of the initial singlet population has been reported,20 which might as well explain the here observed timeconstants. Another explanation for the fast decay of the TRSHG signal might be the geminate recombination of polaron pairs reported by few groups,34,38 while in contrast others attribute these decay channels to time-constants of hundreds of picoseconds or longer.35,37,39 Regarding the aforementioned ultrafast interchain polaron formation, the question would be, why the TR-SHG signal amplitude decreases by more than 20% for the excitation of singlet excitons on the one hand, while the presence of interchain polarons does not affect the signal level, on the other hand. Considering the influence of excess charges and unscreened dipoles on the nonlinear susceptibility of surrounding molecules revealed in theoretical studies and hyper Raman scattering experiments,40,41 an alternative explanation for the observed dynamics appears to be more likely. It is known that the initial hot exciton population is highly delocalized and that their thermalization is accompanied by localization on ultrafast time scales.9,36,42,43 This delocalization can consequently be assumed to induce pronounced distortions in the electron density. As the latter correlates with the nonlinear susceptibility, the observed TR-SHG signal evolution is assigned to originate from the initial excitation and subsequent energetic relaxation and corresponding localization of hot singlet excitons. To elucidate why the thermalization within the donor occurs with two time-constants, the sample is investigated by means of spectroscopic SHG and UV/vis absorption measurements. Figure 3A shows the TR-SHG data of pristine P3HT obtained

Figure 4. TR-SHG data of pristine C60. The initial excitation (crosscorrelation shown in red) results in an increasing TR-SHG signal level, which remains constant for at least 20 ps (cf. Supporting Information, Figure S2). As this wavelength is insufficient for the excitation of triplet excitonic states and singlet transitions are optically dipole forbidden, the change in the NLO response is attributed to the population of long-living charged polarons. Figure 3. Comparing spectroscopic TR-SHG data on pristine P3HT obtained by varying the pump beam wavelength from 579 to 652 nm (A) to the absorbance of the sample (B) elucidates the origin of the two relaxation channels. At 652 nm pump wavelength solely the S1 band is populated and only the fast decay is observed (τ1 = 160 ± 50 fs). At higher excitation energies also the S2 band gets populated and the slow relaxation channel is additionally observed. The black dashed lines in panel B indicate the (vibronic) absorption bands S1 to S3. The colored bars depict the applied excitation wavelengths.

shown in red. The SHG signal remains at the high level within the 20 ps time window of our measurement (see Supporting Information, Figure S2), indicating the population of a longliving excitonic state. The optically dipole allowed direct excitation of triplet excitons can be excluded as the origin of the SHG signal change due to the insufficient applied pump photon energy.29 In contrast, Frenkel-type excitonic states with nanosecond lifetimes were reported (although dipole forbidden) to be optically excitable with photon energies down to 1.57 eV.28,29 However, in this case the energetic relaxation within the singlet manifold would apparently not be accompanied by localization as obtained for the polymer chains. An alternative explanation for the observed NLO contrast, which accounts for the electron affinity of the acceptor molecules, could be the direct excitation of charged polarons as predicted by theory48,49 and observed in time-resolved

for different excitation wavelengths. In contrast to the biexponential decay found using pump wavelengths of 576 and 618 nm, only the fast decay is observed subsequently to the 652 nm excitation, yielding τ = 160 ± 50 fs. Figure 3B depicts the corresponding absorbance spectrum of the sample, inhering the (vibronic) absorption bands S1 to S3, which derive from the 28529



Article

spectroscopy.50 Another argument against the Frenkel-type singlet excitons besides the forbidden optical transition is the expected formation of band-like states in fullerene aggregates.14 For these a change in localization on ultrafast time scales as observed for the semiconducting polymer and an according dynamic alteration of the NLO response would be expected. The charged polarons, however, constitute pairs of charges on adjacent molecules with a constant degree of localization and long-term stability.50 In this case, the symmetry breaking interface would be expected to give rise to an electric field enhancement of the SHG signal. Therefore, the latter assignment appears to be the more reasonable interpretation for the observed change in the SHG signal amplitude. Noticeably, the overall NLO response of the C60 sample is more than 1 order of magnitude smaller than for the pristine P3HT, while the pump-induced relative changes are of comparable magnitude. On the one hand, this corresponds to the expectations as the applied visible wavelength is scarcely absorbed by the acceptor material. On the other hand, this circumstance is also necessary for the unambiguous interpretation of the data on the D/A sample as derived hereinafter. Figure 5A shows the TR-SHG data of the C60/P3HT sample. The red dashed curve is given as a guide to the eye. It

evolution of the signal reflects a completely different behavior, as the SHG signal amplitude does not return to the initial level. It remains constant at the low level for the 50 ps time window of our measurement (see Supporting Information, Figure S3). It is important to note that the signal level at large delays cannot originate from a state that is directly populated by the pump beam as was the case for pristine C60. This is qualitatively illustrated by the red dash−dotted line given as a guide to the eye, which is obtained by multiplying the normalized C60 signal by a factor of −0.5. As the topmost C60 layer is only 12 nm thin and the 579 nm excitation wavelength is within the optical gap of the acceptor,29 the pump beam initially predominantly excites hot, delocalized singlet excitons inside the donating P3HT layer. In the presence of the electron acceptor, an additional interface-derived de-excitation pathway besides the fast thermalization channel within P3HT exists, namely, the charge transfer to C60. Hence, the local maximum in the SHG signal of the D/A sample observed at a pumpprobe delay of approximately 350 fs denotes the turning point between the two nonequilibrium contributions to the TR-SHG signal, viz.: (1) the decreasing distortion of the interfacial electron density by hot delocalized excitons (Figure 5B, Δt > 0) and (2) the field enhancement from the charge transfer across the interface (Figure 5B, Δt ≫ 0). Note that the “field enhancement” results in a decrease of the TR-SHG signal amplitude as the signal change ΔICT is proportional to the electric field (vide inf ra), which in the present case points out of the sample in contrast to the more commonly chosen opposite geometry.8,10,51 The green curve in Figure 5A results from the iteratively convoluting fit, which incorporates the thermalization of hot excitons and the delayed population of the CT state out of the initially excited singlet excitons (for computational details see Supporting Information). It yields τP3HT = 146 ± 8 fs, identical to the localization within P3HT, and τCT = 320 ± 20 fs for the population of the CT state. The latter time-constant fits well to other reported subpicosecond population times of CT states in similar systems as P3HT:PCBM,37 P3HT blends and bulk-heterojunctions with silicon,20 PCPDTBT:PCBM,11 or copper-phthalocyanine bilayers with C60 and C70.10 The absence of recombination, judged from the constantly high CT-induced EFISH signal level on a picosecond time scale found for the latter systems, is in close correlation to the here observed long-living CT state. Analogously, an almost constant photoinduced absorption signal on time scales of hundreds of picoseconds in a polymer:PCBM blend originating from residual holes in the donor were found by Gélinas et al. in time-resolved transient absorption measurements.14 Interestingly, a further population of the CT state delayed by several picoseconds due to exciton drift as reported in the latter study and also by Guo et al.37 is not observed here (cf. Supporting Information, Figure S3). Apparently the majority of the charge transfer process occurs on the ultrafast time scales prior to the hot exciton thermalization and localization. An explanation for the subordinate role of exciton diffusion for the observed CT might be an immobilization of electrons at defect sites,52 at impurities,53 or possibly in less aggregated polymer regions exhibiting slow and dispersive charge transport,25 which emphasizes the importance of the film morphology and accordingly of sample preparation parameters in bulkheterojunction OPVs.3,54 Furthermore, the picosecond decay channel assigned to the thermalization and localization within

Figure 5. (A) TR-SHG data of C60/P3HT differs significantly from the sum of the pristine sample data (red dashed line). The low signal level at large delays originates from the CT induced field enhancement. The red dash−dotted line is a guide to the eye to indicate that the CT state is not directly populated by the pump beam. (B) Δt < 0: SHG probes the steady-state χ(2)-tensor. Δt > 0: Distortion in the interfacial electron density due to highly delocalized hot excitons. Δt ≫ 0: Energetic relaxation and localization of hot excitons and simultaneous electric field formation due to the CT.

constitutes the sum of the TR-SHG data of both pristine samples (cf. Figures 2 and 4), which should reflect the expected qualitative observation if the presence of the acceptor would not influence the induced dynamics in the donor and vice versa. While the initial signal decrease at short delays resembles the dynamics observed for pristine P3HT, the long-term temporal 28530



Article

the S2 band and the subsequent S2 → S1 interband relaxation observed for the pristine P3HT sample is not observed. As it appears the charge transfer becomes the most efficient relaxation channel upon excitation of the S2 band. This finding allows for the important conclusion that a strong coupling between the S2 band and the CT state exists, which could be exploitable to increase the charge transfer yield by providing excess excitation energy, as elucidated in the following. The amount of transferred charge is approximately proportional to the change in the SHG signal level at large delays:51 ΔICT ≈ ηI2(ω)EDC, where the proportionality factor η contains the dependency on the dielectric function and the nonlinear susceptibility,55 I(ω) is the probe beam intensity, and EDC ∝ QCT is the interfacial electric field, which is related to the transferred charge in the approximation of a parallel plate capacitor. Accordingly, the correlation between the CT yield and the excess excitation energy can be elucidated by measuring the SHG signal amplitude change ΔICT(λ) at large delays as a function of the pump photon energy. However, as the absorbance of the donor also depends on the excitation wavelength, ΔICT(λ) needs to be normalized to the number of initially excited electronhole pairs. In the case of sufficiently low excitation densities to avoid undesired effects as exciton exciton annihilation3,9,30−32 or even thermally induced absorbance shifts33 (and for small initial TR-SHG signal amplitude changes ΔI0(λ) in comparison to the steady state signal level, see Supporting Information), ΔI0(λ) is proportional to the initial exciton population. This is indeed the case for the conducted experiments as depicted in Figure 6. Both SHG signal changes at large and short delays ΔICT(λ) and ΔI0(λ), respectively, scale linearly with the pump beam intensity. If the excitons would act as quenchers for each other, the SHG signal changes as a function of the pump intensity would increasingly saturate. Therefore, ΔICT(λ)/ ΔI0(λ) is constant as a function of the pump power (Figure 6B blue curve), while as a function of the excitation wavelength, this term constitutes a measure for the CT yield. In a next step, the photon energy of the pump beam is varied to excite the polymer energetically above (576 nm), in overlap with (618 nm), and below (652 nm), the S0 → S1 absorption maximum centered at 610 nm (cf. Figure 3B). Figure 7A shows the spectroscopic TR-SHG data of the C60/P3HT sample. The dynamics induced by the intermediate photon energy are identical (τP3HT = 146 ± 7 fs, τCT = 320 ± 20 fs) as for the 579 nm excitation. For the 652 nm excitation the charge transfer is slightly faster (τP3HT = 154 ± 26 fs, τCT = 249 ± 66 fs). From these results two important conclusions can be derived. First, the charge transfer is also operative upon the 652 nm excitation although the S2 band of the polymer is not populated. In conjunction with the above finding of the direct, ultrafast CT population out of the S2 band for higher excitation energies, the according conclusion is depicted in Figure 7B: the CT state gets populated out of the S1 band of the polymer, and the process is enhanced when the excess excitation energy is sufficient to populate the S2 band in addition. Second, this results in an increased CT yield as a function of the excitation wavelength as shown in Figure 7C. The charge transfer yield increases by approximately 50% when the pump photon energy is altered from 1.90 to 2.14 eV. These numbers do not only confirm the qualitatively evident competition between energetic relaxation and localization of hot excitons and the charge transfer (cf. Figure 5). They also show that the balance between these two channels progressively shifts in favor of the CT state

Figure 6. (A) TR-SHG data of C60/P3HT measured for varying excitation powers between 10 and 80 μJ/cm2. (B) The SHG signal amplitude changes at short and large delays ΔI0 (red) and ΔICT (black), respectively, scale linearly with the excitation power. ΔICT/ΔI0 (blue), therefore, constitutes a measure for the relative amount of transferred charge. The minor observed offsets are a result of the direct excitation of the topmost acceptor layer (cf. Supporting Information, Figure S4).

population with increasing excess energy at the higher energetic excitations.

4. CONCLUSIONS These results that reveal an efficient CT state population out of initially excited hot singlet excitons are in agreement with the recently reported excitation of hot CT states enabled when using pump beams with high photon energies.11 Albeit, in the present study we solely focused on the charge transfer as one essential step in the light-to-current conversion in OPVs and not also on the subsequent step, the charge separation, as Grancini et al. did.11 Of course, it has to be kept in mind that an increased charge transfer yield does not necessarily lead to an increase of the charge generation yield. Recent studies, for example, revealed no correlation between the excitation energy and the quantum yield for charge separation in a variety of OPV model systems.15,16 Nevertheless, as demonstrated by Herrmann et al.20 the charge transfer yield drastically depends on the (donor) film morphology and accordingly on sample preparation parameters and material properties as the degree of regioregularity in the case of P3HT. In their study the presence of less aggregated polymer domains inhering a blue-shifted absorbance in comparison to the π-stacking-mediated selforganized lamella superstructure forming RR-P3HT resulted in the excitation of less delocalized excitons and hence in an inefficient exploitation of the high energetic excitation at short wavelengths. The present study provides two important findings regarding the feasibility of an OPV performance optimization. (1) In well28531



Article

samples and for numerous productive ideas and constructive discussions, respectively.

■

Figure 7. (A) TR-SHG data of C60/P3HT for different excitation wavelengths. (B) The CT state gets populated out of the S1 band of the polymer. The process is enhanced if the excitation energy is sufficient to involve the S2 band in addition. (C) The CT yield ΔICT(λ)/ΔI0(λ) correlates with the excitation wavelength. It rises by approximately 50% when the pump photon energy is increased from 1.90 to 2.14 eV.

defined systems, increased CT yields out of a hot exciton population are realizable. This clearly indicates toward an advantage of low bandgap polymers56−58 aside from considerations regarding the Shockley−Queisser limit. (2) As demonstrated here, combining the inherent interface sensitivity and the femtosecond temporal resolution of the TR-SHG probe with the concept of decomposing OPVs into welldefined, layered model systems enables a stepwise investigation of the light-to-current conversion and hence holds the key to an understanding of the underlying processes in organic photovoltaics.

■

ASSOCIATED CONTENT

S Supporting Information *

Details on sample preparation, the experimental setup, data processing, and data on polarization-resolved SHG and large delay measurements. This material is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

(1) Dang, M. T.; Hirsch, L.; Wantz, G. P3HT:PCBM, Best Seller in Polymer Photovoltaic Research. Adv. Mater. 2011, 23, 3597−3602. (2) Chen, L.-M.; Hong, Z.; Li, G.; Yang, Y. Recent Progress in Polymer Solar Cells: Manipulation of Polymer:Fullerene Morphology and the Formation of Efficient Inverted Polymer Solar Cells. Adv. Mater. 2009, 21, 1434−1449. (3) Shaw, P. E.; Ruseckas, A.; Samuel, I. D. W. Exciton Diffusion Measurements in Poly(3-hexylthiophene). Adv. Mater. 2008, 20, 3516−3520. (4) Kim, Y.; Cook, S.; Tuladhar, S. M.; Choulis, S. A.; Nelson, J.; Durrant, J. R.; Bradley, D. D. C.; Giles, M.; McCulloch, I.; Ha, C.-S.; et al. Strong Regioregularity Effect in Self-Organizing Conjugated Polymer Films and High-Efficiency Polythiophene:Fullerene Solar Cells. Nat. Mater. 2006, 5, 197−203. (5) Yu, G.; Gao, J.; Hummelen, J. C.; Wudl, F.; Heeger, A. J. Polymer Photovoltaic Cells: Enhanced Efficiencies via a Network of Internal Donor-Acceptor Heterojunctions. Science 1995, 270, 1789−1791. (6) Dennler, G.; Scharber, M. C.; Brabec, C. J. Polymer-Fullerene Bulk-Heterojunction Solar Cells. Adv. Mater. 2009, 21, 1323−1338. (7) Green, M. A.; Emery, K.; Hishikawa, Y.; Warta, W.; Dunlop, E. D. Solar Cell Efficiency Tables (Version 39). Prog. Photovolt. Res. Appl. 2012, 20, 12−20. (8) Chan, W.-L.; Ligges, M.; Jailaubekov, A.; Kaake, L.; Miaja-Avila, L.; Zhu, X.-Y. Observing the Multiexciton State in Singlet Fission and Ensuing Ultrafast Multielectron Transfer. Science 2011, 334, 1541− 1545. (9) Chen, K.; Barker, A. J.; Reish, M. E.; Gordon, K. C.; Hodgkiss, J. M. Broadband Ultrafast Photoluminescence Spectroscopy Resolves Charge Photogeneration via Delocalized Hot Excitons in Polymer:Fullerene Photovoltaic Blends. J. Am. Chem. Soc. 2013, 135, 18502− 18512. (10) Jailaubekov, A. E.; Willard, A. P.; Tritsch, J. R.; Chan, W.-L.; Sai, N.; Gearba, R.; Kaake, L. G.; Williams, K. J.; Leung, K.; Rossky, P. J.; et al. Hot Charge-Transfer Excitons Set the Time Limit for Charge Separation at Donor/Acceptor Interfaces in Organic Photovoltaics. Nat. Mater. 2013, 12, 66−73. (11) Grancini, G.; Maiuri, M.; Fazzi, D.; Petrozza, A.; Egelhaaf, H.-J.; Brida, D.; Gerullo, G.; Lanzani, G. Hot Exciton Dissociation in Polymer Solar Cells. Nat. Mater. 2013, 12, 29−33. (12) Borges, I., Jr.; Aquino, A. J. A.; Köhn, A.; Nieman, R.; Hase, W. L.; Chen, L. X.; Lischka, H. Ab Initio Modeling of Excitonic and Charge-Transfer States in Organic Semiconductors: The PTB1/ PCBM Low Band Gap System. J. Am. Chem. Soc. 2013, 135, 18252− 18255. (13) Onsager, L. Initial Recombination of Ions. Phys. Rev. 1938, 54, 554−557. (14) Gélinas, S.; Rao, A.; Kumar, A.; Smith, S. L.; Chin, A. W.; Clark, J.; van der Poll, T. S.; Bazan, G. C.; Friend, R. H. Ultrafast Long-Range Charge Separation in Organic Semiconductor Photovoltaic Diodes. Science 2014, 343, 512−516. (15) Vandewal, K.; Albrecht, S.; Hoke, E. T.; Graham, K. R.; Widmer, J.; Douglas, J. D.; Schubert, M.; Mateker, W. R.; Bloking, J. T.; Burkhard, G. F.; et al. Efficient Charge Generation by Relaxed ChargeTransfer States at Organic Interfaces. Nat. Mater. 2014, 13, 63−68. (16) Albrecht, S.; Vandewal, K.; Tumbleston, J. R.; Fischer, F. S. U.; Douglas, J. D.; Fréchet, J. M. J.; Ludwigs, S.; Ade, H.; Salleo, A.; Neher, D. On the Efficiency of Charge Transfer State Splitting in Polymer:Fullerene Solar Cells. Adv. Mater. 2014, 26, 2533−2539. (17) McGilp, J. F. A Review of Optical Second-Harmonic and SumFrequency Generation at Surfaces and Interfaces. J. Phys. D: Appl. Phys. 1996, 29, 1812−1821. (18) Shen, Y. R. Surface Properties Probed by Second-Harmonic and Sum-Frequency Generation. Nature 1989, 337, 519−525.

AUTHOR INFORMATION

Corresponding Authors

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

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Funding by the Deutsche Forschungsgemeinschaft (DFG) through the priority program SPP 1355 is gratefully acknowledged. The authors thank Steffen Roland, Marcel Schubert, and Dieter Neher (Universität Potsdam) for preparation of the 28532



Article

(19) Bouhelier, A.; Beversluis, M.; Hartschuh, A.; Novotny, L. NearField Second-Harmonic Generation Induced by Local Field Enhancement. Phys. Rev. Lett. 2003, 90 (013903), 1−4. (20) Herrmann, D.; Niesar, S.; Scharsich, C.; Köhler, A.; Stutzmann, M.; Riedle, E. Role of Structural Order and Excess Energy on Ultrafast Free Charge Generation in Hybrid Polythiophene/Si Photovoltaics Probed in Real Time by Near-Infrared Broadband Transient Absorption. J. Am. Chem. Soc. 2011, 133, 18220−18233. (21) Ma, W.; Yang, C.; Gong, X.; Lee, K.; Heeger, A. J. Thermally Stable, Efficient Polymer Solar Cells with Nanoscale Control of the Interpenetrating Network Morphology. Adv. Funct. Mater. 2005, 15, 1617−1622. (22) Campoy-Quiles, M.; Ferenczi, T.; Agostinelli, T.; Etchegoin, P. G.; Kim, Y.; Anthopoulos, T. D.; Stavrinou, P. N.; Bradley, D. D. C.; Nelson, J. Morphology Evolution via Self-Organization and Lateral and Vertical Diffusion in Polymer:Fullerene Solar Cell Blends. Nat. Mater. 2008, 7, 158−164. (23) Xu, Z.; Chen, L.-M.; Chen, M.-H.; Li, G.; Yang, Y. Energy Level Alignment of Poly(3-hexylthiophene): [6,6]-Phenyl C61 Butyric Acid Methyl Ester Bulk Heterojunction. Appl. Phys. Lett. 2009, 95 (013301), 1−3. (24) Agostinelli, T.; Lilliu, S.; Labram, J. C.; Campoy-Quilles, M.; Hampton, M.; Pires, E.; Rawle, J.; Bikondoa, O.; Bradley, D. D. C.; Anthopoulos, D.; et al. Real-Time Investigation of Crystallization and Phase-Segregation Dynamics in P3HT:PCBM Solar Cells During Thermal Annealing. Adv. Funct. Mater. 2011, 21, 1701−1708. (25) Mauer, R.; Kastler, M.; Laquai, F. The Impact of Polymer Regioregularity on Charge Transport and Efficiency of P3HT:PCBM Photovoltaic Devices. Adv. Funct. Mater. 2010, 20, 2085−2092. (26) Chirvase, D.; Parisi, J.; Hummelen, J. C.; Dyakonov, V. Influence of Nanomorphology on the Photovoltaic Action of Polymer−Fullerene Composites. Nanotechnology 2004, 15, 1317− 1323. (27) Jiang, X. M.; Ö sterbacka, R.; An, C. P.; Vardeny, Z. V. Photoexcitations in Regio-regular and Regio-random Polythiophene Films. Synth. Met. 2003, 137, 1465−1468. (28) Kuhnke, K.; Becker, R.; Kern, K. Dynamics of Second Harmonic Generation at the C60/Quartz Interface. Chem. Phys. Lett. 1996, 257, 569−575. (29) Janner, A.-M.; Eder, R.; Koopmans, R.; Jonkman, H. T.; Sawatzky, G. A. Excitons in C60 Studied by Temperature-Dependent Optical Second-Harmonic Generation. Phys. Rev. B 1995, 52, 17158− 17164. (30) Nguyen, T.-Q.; Martini, I. B.; Liu, J.; Schwartz, B. J. Controlling Interchain Interactions in Conjugated Polymers: The Effects of Chain Morphology on Exciton−Exciton Annihilation and Aggregation in MEH−PPV Films. J. Phys. Chem. B 2000, 104, 237−255. (31) Lewis, A. J.; Ruseckas, A.; Gaudin, O. P. M.; Webster, G. R.; Burn, P. L.; Samuel, I. D. W. Singlet Exciton Diffusion in MEH-PPV Films Studied by Exciton−Exciton Annihilation. Org. Electron. 2006, 7, 452−456. (32) Silva, C.; Dhoot, A. S.; Russell, D. M.; Stevens, M. A.; Arias, A. C.; MacKenzie, J. D.; Greenham, N. C.; Friend, R. H.; Setayesh, S.; Müllen, K. Efficient Exciton Dissociation via Two-Step Photoexcitation in Polymeric Semiconductors. Phys. Rev. B 2001, 64 (125211), 1−7. (33) Albert-Seifried, S.; Friend, R. H. Measurement of Thermal Modulation of Optical Absorption in Pump-Probe Spectroscopy of Semiconducting Polymers. Appl. Phys. Lett. 2011, 98 (223304), 1−3. (34) Korovyanko, O. J.; Ö sterbacka, R.; Jiang, X. M.; Vardeny, Z. V. Photoexcitation Dynamics in Regioregular and Regiorandom Polythiophene Films. Phys. Rev. B 2001, 64 (235122), 1−6. (35) Piris, J.; Dykstra, T. E.; Bakulin, A. A.; van Loosdrecht, P. H. M.; Knulst, W.; Trinh, M. T.; Schins, J. M.; Siebbeles, L. D. A. Photogeneration and Ultrafast Dynamics of Excitons and Charges in P3HT/PCBM Blends. J. Phys. Chem. C 2009, 113, 14500−14506. (36) Guo, J.; Ohkita, H.; Benten, H.; Ito, S. Near-IR Femtosecond Transient Absorption Spectroscopy of Ultrafast Polaron and Triplet

Exciton Formation in Polythiophene Films with Different Regioregularities. J. Am. Chem. Soc. 2009, 131, 16869−16880. (37) Guo, J.; Ohkita, H.; Benten, H.; Ito, S. Charge Generation and Recombination Dynamics in Poly(3-hexylthiophene)/Fullerene Blend Films with Different Regioregularities and Morphologies. J. Am. Chem. Soc. 2010, 132, 6154−6164. (38) Kandada, A. R. S.; Grancini, G.; Petrozza, A.; Perissinotto, S.; Fazzi, D.; Raavi, S. S. K.; Lanzani, G. Ultrafast Energy Transfer in Ultrathin Organic Donor/Acceptor Blend. Sci. Rep. 2013, 3 (2073), 1−7. (39) Dogariu, A.; Vacar, D.; Heeger, A. J. Picosecond Time-Resolved Spectroscopy of the Excited State in a Soluble Derivative of Poly(phenylene vinylene): Origin of the Bimolecular Decay. Phys. Rev. B 1998, 58, 10218−10224. (40) Terenziani, F.; Ghosh, S.; Robin, A.-C.; Das, P. K.; BlanchardDesce, M. Environmental and Excitonic Effects on the First Hyperpolarizability of Polar Molecules and Related Dimers. J. Phys. Chem. B 2008, 112, 11498−11505. (41) Datta, A.; Pati, S. K. Effects of Dipole Orientations on Nonlinear Optical Properties of Oxo-Bridged Dinitroaniline Systems. J. Phys. Chem. A 2004, 108, 320−325. (42) Ruseckas, A.; Wood, P.; Samuel, I. D. W.; Webster, G. R.; Mitchell, W. J.; Burn, P. L.; Sundström, V. Ultrafast Depolarization of the Fluorescence in a Conjugated Polymer. Phys. Rev. B 2005, 72 (115214), 1−5. (43) Hwang, I.; Scholes, G. D. Electronic Energy Transfer and Quantum-Coherence in π-Conjugated Polymers. Chem. Mater. 2011, 23, 610−620. (44) Brown, P. J.; Thomas, D. S.; Köhler, A.; Wilson, J. S.; Kim, J.-S.; Ramsdale, C. M.; Sirringhaus, H.; Friend, R. H. Effect of Interchain Interactions on the Absorption and Emission of Poly(3-hexylthiophene). Phys. Rev. B 2003, 67 (064203), 1−16. (45) Clark, J.; Silva, C.; Friend, R. H.; Spano, F. C. Role of Intermolecular Coupling in the Photophysics of Disordered Organic Semiconductors: Aggregate Emission in Regioregular Polythiophene. Phys. Rev. Lett. 2007, 98 (206406), 1−4. (46) Jiang, X. M.; Ö sterbacka, R.; Korovyanko, O.; An, C. P.; Horovitz, B.; Janssen, R. A. J.; Vardeny, Z. V. Spectroscopic Studies of Photoexcitations in Regioregular and Regiorandom Polythiophene Films. Adv. Funct. Mater. 2002, 12, 587−597. (47) Spano, F. C. Modeling Disorder in Polymer Aggregates: The Optical Spectroscopy of Regioregular Poly(3-hexylthiophene) Thin Films. J. Chem. Phys. 2005, 122 (234701), 1−15. (48) Harigaya, K. Lattice Distortion and Energy-level Structures in Doped C60 and C70 Molecules Studied with the Extended SuSchrieffer-Heeger Model: Polaron Excitations and Optical Absorption. Phys. Rev. B 1992, 45, 13676−13684. (49) Friedman, B. Electronic Absorption Spectra in C60− and C60+. Phys. Rev. B 1993, 48, 2743−2747. (50) Dick, D.; Wei, X.; Jeglinski, S.; Benner, R. E.; Vardeny, Z. V.; Moses, D.; Srdanov, V. I.; Wudl, F. Transient Spectroscopy of Excitons and Polarons in C60 Films from Femtoseconds to Milliseconds. Phys. Rev. Lett. 1994, 73, 2760−2763. (51) Tisdale, W. A.; Williams, K. J.; Timp, B. A.; Norris, D. J.; Aydil, E. S.; Zhu, X.-Y. Hot-Electron Transfer from Semiconductor Nanocrystals. Science 2010, 328, 1543−1547. (52) Graupner, W.; Leditzky, G.; Leising, G.; Scherf, U. Shallow and Deep Traps in Conjugated Polymers of High Intrachain Order. Phys. Rev. B 1996, 54, 7610−7613. (53) Nicolai, H. T.; Kuik, M.; Wetzelaer, G. A. H.; de Boer, B.; Campbell, C.; Risko, C.; Brédas, J. L.; Blom, P. W. M. Unification of Trap-Limited Electron Transport in Semiconducting Polymers. Nat. Mater. 2011, 11, 882−887. (54) Drummy, L. F.; Davis, R. J.; Moore, D. L.; Durstock, M.; Vaia, R. A.; Hsu, J. W. P. Molecular-Scale and Nanoscale Morphology of P3HT:PCBM Bulk Heterojunctions: Energy-Filtered TEM and LowDose HREM. Chem. Mater. 2011, 23, 907−912. 28533



Article

(55) Nahata, A.; Heinz, T. F. Detection of Freely Propagating Terahertz Radiation by Use of Optical Second-Harmonic Generation. Opt. Lett. 1998, 23, 67−69. (56) Peet, J.; Kim, J. Y.; Coates, N. E.; Ma, W. L.; Moses, D.; Heeger, A. F.; Bazan, G. C. Efficiency Enhancement in Low-Bandgap Polymer Solar Cells by Processing with Alkane Dithiols. Nat. Mater. 2007, 6, 497−500. (57) Grancini, G.; Martino, N.; Antognazza, M. R.; Celebrano, M.; Egelhaaf, H.-J.; Lanzani, G. Influence of Blend Composition on Ultrafast Charge Generation and Recombination Dynamics in Low Band Gap Polymer-Based Organic Photovoltaics. J. Phys. Chem. C 2012, 116, 9838−9844. (58) Cowan, S. R.; Banerji, N.; Leong, W. L.; Heeger, A. J. Charge Formation, Recombination, and Sweep-Out Dynamics in Organic Solar Cells. Adv. Funct. Mater. 2012, 22, 1116−1128.

28534


Acceptor Charge Transfer Yield - The

Recommend Documents