Transient absorption spectroscopy of charge photogeneration yields and lifetimes in a low bandgap polymer /fullerenefilm

Abstract

Charge photogeneration yields and energetics are determined for a low band gap co-polymer , PCPDTBT blended with PC₇₀BM; the increase in charge photogeneration with dithiol is correlated with an increase in the free energy of charge separation.

---

Attempts to increase the efficiency of organic photovoltaic cells beyond the ~4% produced by poly(3-hexylthiophene) (P3HT) blended with 6,6-phenyl C₆₁-butyric acid methyl ester (PC₆₀BM)¹ often involve the synthesis of polymers with smaller optical band gaps.^2,3 This provides an improved spectral overlap with the solar irradiation and thus enhances the light absorption. However this smaller band gap reduces the energy of photogenerated excitons , thereby potentially reducing the free energy available to drive charge separation.

We have recently reported a study of polaron yields for a series of polythiophenes blended with PCBM. A reduction of the effective free energy of charge separation, ΔG_CS^rel by 300 meV was observed to result in a decrease in the yield of dissociated polarons by two orders of magnitude, indicating that for this class of polymers a relatively large ΔG_CS^rel is required to achieve efficient charge photogeneration.⁴ Contrasting with this observation, it has recently been reported that the low band gap co-polymer , PCPDTBT, poly[2,6-(4,4-bis(2-ethylhexyl)-4H-cyclopenta[2,1-b;3,4-b′]dithiophene)-alt-4,7-(2,1,3-benzothiadiazole], when blended with PC₇₀BM (see Scheme 1), can result in remarkably efficient photovoltaic device performance. By the addition of an alkanedithiol (a processing additive) to the solvent prior to spin coating, device efficiencies of over 5% have been reported.⁵ These observations raise the possibility that charge photogeneration in blend films based on the PCPDTBT co-polymer may be less dependent upon a large ΔG_CS^rel than films based on the polythiophene series studied previously. In this paper we therefore employ transient absorption spectroscopy (TAS) to examine the charge photogeneration yields and liftetimes of PCPDTBT–PC₇₀BM (1 : 2) blend films with and without the use of dithiol. These data are quantitatively compared against those we have reported previously for the polythiophene/PCBM film series (which includes P3HT), and discussed in terms of the parameters determining the efficiency of charge photogeneration in such films .

Scheme 1 The molecular structures of PCPDTBT (left) and PC₇₀BM (right).

Typical spin-coated film absorption spectra are shown in Fig. 1. Pristine PCPDTBT exhibits an absorption maximum of 775 nm. The addition of PC₇₀BM is observed to result in a blue-shift of this maximum to 730 nm. This has been attributed to a loss of molecular order in the polymer film , with the PC₇₀BM disrupting the π-stacking and preventing the formation of large aggregated polymer domains⁶ and, as we report elsewhere, reducing the polymer crystallinity. Analogous behaviour has been reported for P3HT/PCBM films .⁷ The addition of 25 mg mL⁻¹1,8-octanedithiol in the original blend solution causes the absorption maximum of the resultant film to red-shift to 790 nm, indicating that the π-stacking lost by the addition of PC₇₀BM has been restored.

Fig. 1 Ground state absorption spectra of the PCPDTBT pristine and blend films and the transient spectrum of a PCPDTBT–PC₇₀BM (1 : 2) film obtained after 1 µs using 30 µJ cm⁻² excitation with a pump wavelength of 650 nm.

Transient absorption spectroscopy of blend films was undertaken as reported previously,⁸ with the use of a In_xGa_1–xAs photodiode detector. A typical transient spectrum of the blend film , measured at 1 µs, is plotted in Fig. 1 (filled squares). This spectrum shows a single maximum at 1280 nm, in agreement with previous transient absorption studies⁹ and is assigned to photoinduced absorption of PCPDTBT polarons. Similarly shaped spectra were observed with and without dithiol, although dithiol did cause the band to narrow slightly, attributed to the enhanced film crystallinity. Control data on pristine PCPDTBT films indicated similar, albeit ten-fold smaller, transients, consistent with the expected relatively low charge photogeneration yield in the pristine film

The decay dynamics of the transient absorption signal observed for the PCPDTBT–PC₇₀BM blend films were measured as a function of excitation density. These decay dynamics were observed to be independent of the presence of oxygen, consistent with our assignment of the transient signals to polarons rather than triplet states. Two phases of recombination dynamics are evident (see inset of Fig. 2): an excitation density-dependent fast phase from 500 ns–10 µs and a slow phase from 10 µs into the ms time regime that is independent of laser excitation density. This biphasic behaviour and dependence upon excitation density is very similar to that we have observed previously for MDMO-PPV–PC₆₀BM blend films ,⁸ except that the fast phase extends into the µs timescale in the case of PCPDTBT. As such, the slow phase, which can be fit to a power law decay (ΔOD ∝ t^–α, with α = 0.52), is assigned to bimolecular recombination of dissociated polarons in the presence of an exponential tail of intra-band gap localised (trapped) states.¹⁰ This value of α is higher than that observed for blends of MDMO–PPV (α = 0.4), indicative of a relatively low level of deep traps for the PCPDTBT–PC₇₀BM blend films . The fast phase that appears at high laser intensities is assigned to recombination of free polarons that are generated when the density of photo-generated polarons exceeds the density of localised states.

Fig. 2 Transient absorption decays of PCPDTBT–PC₇₀BM (1 : 2) films with and without dithiol obtained using 30 µJ cm⁻² excitation at 650 nm and a probe wavelength of 1300 nm. The inset shows the decays of a film with dithiol as a function of laser excitation density.

Also shown in Fig. 2 is a comparison of transient absorption data with and without the use of the dithiol co-solvent. Similar decay dynamics and dependence upon excitation wavelength were observed for both films . More strikingly, the initial signal amplitude increased approximately two-fold with the inclusion of dithiol to the spin coating solution (we note the amplitude of the slow power law decay phase did not increase, consistent with the saturation of the trap density of states indicated by our studies as function of excitation density). Data were collected at an excitation wavelength (650 nm) where the film ground state absorbance was the same for both films , therefore the increase in signal amplitude cannot be attributed to enhanced light absorption. Furthermore, this increase in signal amplitude was observed for all probe wavelengths measured, indicative of a broad increase in photoinduced PCPDTBT polaron absorption. Whilst a change in polaron optical cross-section cannot be ruled out, the observation of such a general increase in photoinduced absorption is strongly indicative of an enhanced yield of long-lived PCPDTBT polarons.

The two-fold increase in transient absorption signal amplitude with dithiol observed on the microsecond timescale is therefore assigned to an increase in the yield of long-lived dissociated charges. This increase in charge photogeneration does not appear to result from enhanced exciton quenching, with photoluminescence data (see ESI† ) showing similar emission quenching for films deposited with or without dithiol. Rather, as we have discussed previously in the context of polythiophene/PCBM blends,⁴ the observed increase in charge photogeneration can more likely be assigned to a reduction of geminate recombination losses, as we discuss in more detail below.

We have shown previously for polythiophene/PCBM blend films that the yield of dissociated polarons is strongly dependent upon the effective free energy of charge separation, ΔG_C^relf,⁴ defined as the difference between the singlet exciton energy, S₁, of the polymer and the polaron pair energy (estimated from the electron affinity of PCBM subtracted from the ionisation potential of the polymer ): ΔG_CS^rel = S₁ − (IP − EA). The effect of dithiol upon the singlet exciton energy can be readily determined from the absorption spectra shown in Fig. 1, indicating that the addition of dithiol results in a 0.1 eV decrease in singlet exciton energy, attributed to an increase in PCPDTBT crystallinity. In order to estimate the influence of dithiol upon the PCPDTBT ionisation potential, film cyclic voltammetry data were collected, as shown in Fig. 3. From these data, it is apparent that the first oxidation wave of PCPDTBT is shifted positively by 200 meV for the blend film deposited without dithiol compared to either the pristine PCPDTBT film or the blend film deposited with dithiol. This shift is consistent with the changes in absorption spectra and crystallinity discussed above. Such a shift in oxidation potential with polymer crystallinity has also been observed for other polymers , such as cyclic voltammetry studies of P3HT as a function of regioregularity¹¹ and annealing. Taking account of the changes in both S₁ and ionisation (oxidation) potential, we conclude that the addition of dithiol to the spin coating solution results in a net increase in ΔG_CS^rel of ~0.1 eV in the resultant blend films .

Fig. 3 Cyclic voltammograms of a pristine PCPDTBT film and blend films with PC₇₀BM (1 : 2) with and without dithiol, using a scan rate of 10 mV s⁻¹, an Ag/AgCl reference electrode, and a tetrabutylammonium perchlorate/acetonitrile electrolyte.

We now turn to a quantitative comparison of the polaron yields and charge separation energetics we have observed here with those we have observed previously for polythiophene: PCBM blend films . As discussed previously, the yield of dissociated polarons can be approximately estimated from the amplitude of the transient absorption signal at 1 µs, measured at the polaron’s absorption maximum.¹² Details of determination of ΔG_CS^rel for the PCPDTBT relative to P3HT is given in the ESI.† The resultant plots of ΔOD vs. ΔG_CS^rel are shown in Fig. 4. Two points are striking from this plot. Firstly, it is apparent that the influence of dithiol upon the PCPDTBT data, showing a two-fold increase in ΔOD correlated with a 100 meV increase in ΔG_C^relf, is in excellent quantitative agreement with the ΔOD vs. ΔG_CS^rel dependence we observed previously for our polythiophene series. This suggests that the primary origin of the increased polaron yield with dithiol (and thereby enhanced device photocurrent) may indeed be an increase in ΔG_CS^rel caused by increased PCPDTBT crystallinity, which reduces the polymer’s ionisation potential. Alternatively, this enhanced polaron yield could be due to an increase in the PCBM domain size, as we will discuss elsewhere. Secondly, it is apparent that both PCPDTBT blend films exhibit ΔOD signal amplitudes over two orders of magnitude greater than those expected for polythiophene–PCBM films with comparable ΔG_CS^rel. This strongly suggests that less energetic driving force is required to generate dissociated charges in PCPDTBT blend films compared to those of polythiophenes.

Fig. 4 Plot of the ΔOD signal amplitude vs. ΔG_CS^rel for PCPDTBT–PC₇₀BM (1 : 2) blend films deposited with or without dithiol. Also shown are the corresponding data for various 19 : 1 polythiophene blend films with PC₆₀BM, reproduced from ref. 5. Note that for P3HT, similar data were obtained for both 19 : 1 and 1 : 2 blend films and for PC₇₀BM and PC₆₀BM acceptors, indicating that the differences between the data points shown in this graph can be primarily attributed to the properties of the polymers employed. Note that ΔG_CS^rel should be regarded only as a relative rather than absolute measure of interfacial energetics.

We have previously proposed that the strong dependence of dissociated polaron generation yield upon ΔG_CS^rel may result from the dissociation of the initially formed bound (interfacial) radical pairs requiring excess thermal energy to overcome their coulomb attraction. In this model, the low charge generation yields observed for polythiophene/PCBM blends with low ΔG_C^relf is assigned to monomolecular (geminate) recombination. In this context, the high polaron yields observed for the PCPDTBT–PCBM films may be associated with the partial charge transfer nature of the polymer , with the LUMO orbital being localised on the benzothiadiazole moiety. Such charge transfer or ‘redox relay’ motifs are well known to facilitate charge dissociation in other systems, including dye sensitized solar cells. Whatever the origin, it appears that the ability of PCPDTBT : PCBM blend films to achieve charge separation with only a small energetic driving force is likely to be a key reason behind the remarkably high photovoltaic device efficiencies achieved with such blends.

We thank Konarka and the EPSRC for funding, and Toby Ferenczi and James Kirkpatrick for helpful discussions, and the latter for assistance with the front cover illustration.

Notes and references

1. W. W. Ma, C. Yang, X. Gong, K. Lee and A. J. Heeger, Adv. Funct. Mater., 2005, 15, 1617.
2. M. C. Scharber, D. Mühlbacher, M. Koppe, P. Denk, C. Waldauf, A. J. Heeger and C. J. Brabec, Adv. Mater. (Weinheim, Ger.), 2006, 18, 789–794.
3. C. Soci, I. W. Hwang, D. Moses, Z. Zhu, D. Waller, R. Gaudiana, C. J. Brabec and A. J. Heeger, Adv. Funct. Mater., 2007, 17, 632–636.
4. H. Okhita, S. Cook, Y. Astuti, W. Duffy, S. Tierney, W. Zhang, M. Heeney, I. McCulloch, J. Nelson, D. D. C. Bradley and J. R. Durrant, J. Am. Chem. Soc., 2008, 130, 3030–3042.
5. J. Peet, J. Y. Kim, N. E. Coates, W. L. Ma, D. Moses, A. J. Heeger and G. C. Bazan, Nat. Mater., 2007, 6, 497–500.
6. M. Morana, M. Wegscheider, A. Bonanni, N. Kopidakis, S. Shaheen, M. Scharber, Z. Zhu, D. Waller, R. Gaudiana and C. Brabec, Adv. Funct. Mater., 2008, 18, 1757–1766.
7. T. Erb, U. Zhokhavets, G. Gobsch, S. Raleva, B. Stühn, P. Schilinsky, C. Waldauf and C. J. Brabec, Adv. Funct. Mater., 2005, 15, 1193–1196.
8. A. F. Nogueira, I. Montanari, J. Nelson, J. R. Durrant, C. Winder, N. S. Sariciftci and C. Brabec, J. Phys. Chem. B, 2003, 107, 1567–1573.
9. I.-W. Hwang, C. Soci, D. Moses, Z. Zhu, D. Waller, R. Gaudiana, C. J. Brabec and A. J. Heeger, Adv. Mater. (Weinheim, Ger.), 2007, 19, 2307–2312.
10. J. Nelson, Phys. Rev. B: Condens. Matter Mater. Phys., 2003, 67, 155209.
11. M. M. Skompska and A. Szkurłat, Electrochim. Acta, 2001, 46, 4007–4015.
12. The magnitude of the observed variation in ΔOD is too large to result only from other possible determinants, such as variations in polaron absorption coefficient.

---

Footnote

† Electronic supplementary information (ESI) available: PL quenching data for PCPDTBT (pristine, blends with/without dithiol), ΔG calculation for PCPDTBT. See DOI: 10.1039/b813815j

---