Temperature/time-dependent crystallization of polythiophene:fullerene bulk heterojunction films for polymer solar cells

Abstract

We report the temperature/time-dependent crystallization of poly(3-hexylthiophene) (P3HT) in blend films of P3HT and [6,6]-phenyl-C₆₁-butyric acid methyl ester (PC₆₁BM). The crystallization behaviour of P3HT:PC₆₁BM blend films was measured as a function of annealing time at two different temperatures (150 °C and 160 °C) by employing a synchrotron-radiation grazing-incidence angle X-ray diffraction (GIXD) technique. The crystallization behaviour was correlated with corresponding solar cells annealed under the same conditions. Results showed that the trend of device performance was almost in accordance with that of the (100) GIXD intensity, indicating that the nanostructure change in blend films does affect the device performance. However, the intermediate zones related to nanomorphology fluctuations, which were observed for lower temperature (140 °C) annealing, were significantly suppressed at higher temperature (150 °C and 160 °C) annealing.

---

Introduction

Polymer:fullerene solar cells have recently attracted a great deal of attention due to their intrinsic advantages including low-cost manufacturing and a variety of features such as flexibility, light weight, ultra-thin shape, etc.^1–4 The landmark breakthroughs in polymer:fullerene solar cells can be attributed to solvent-induced nanomorphology control in poly[2-methoxy-5-(3′,7′-dimethyloctyloxy)-1,4-phenylene vinylene] (MDMO-PPV):[6,6]-phenyl-C₆₁-butyric acid methyl ester (PC₆₁BM) blend films⁵ and to the thermal and/or solvent annealing in poly(3-hexylthiophene) (P3HT):PC₆₁BM blend films,^6–15 even though further improvement has been made by employing newly synthesized conjugated polymers and [6,6]-phenyl-C₇₁-butyric acid methyl ester (PC₇₁BM).^16,17

Of the various kinds of polymer:fullerene solar cells, to date, P3HT:PCBM solar cells have been extensively studied and have become a standard device for polymer solar cell research because their performance is reproducible with known fabrication conditions. This outstanding reproducibility can be ascribed to the self-assembly features of P3HT which drive quick crystallization upon thermal and/or solvent annealing in the presence of regioregularity dependency.^10–15 In addition, it has been reported that the formation of particular fullerene crystal nanostructures is also responsible for guaranteeing high power-conversion efficiencies in P3HT:PC₆₁BM solar cells.¹⁸

Recently we found that the nanomorphology of P3HT:PC₆₁BM solar cells is sensitively dependent on the annealing time at a fixed annealing temperature.¹⁹ In particular, the initial stability of P3HT:PC₆₁BM solar cells was found to be significantly influenced by the annealing time.²⁰ However, there remains a question about the correlation between annealing time and temperature because the previous work revealed the time-dependent morphology fluctuations at only a single temperature.¹⁸

In this work, we investigate the time-dependent crystallization of P3HT:PC₆₁BM blend films at two different temperatures in order to understand the correlation between annealing time and temperature. As a probe for monitoring the crystallization (nanostructure changes), we employed a synchrotron-radiation grazing-incidence angle X-ray diffraction (GIXD) technique. The trend of time-dependent crystallization in the blend films was correlated with that of time-dependent device performances. Results showed that the time-dependent crystallization behaviour of P3HT phases in the blend films was affected by annealing temperatures. In particular, the device performance was found to be in good agreement with the GIXD intensity change irrespective of annealing temperature.

Experimental section

Materials and solution preparation

P3HT (weight-average molecular weight = 5.9 × 10⁴; polydispersity index = 2.32; regioregularity = 92.2%) and PC₆₁BM (purity = 99.5%) were used without further purification as received from Rieke Metals and Nano-C, respectively. The P3HT:PC₆₁BM blend solutions were prepared using chlorobenzene as a solvent: The solid concentration was 60 mg ml⁻¹, while the weight ratio of P3HT to PC₆₁BM was 1 : 1. These solutions were continuously mixed prior to spin-coating.

Blend film and device fabrication

To make blend films and devices, indium-tin oxide (ITO)-coated glass substrates were patterned using a photolithography technique: The ITO stripe was 3 mm wide and 12 mm long. These patterned ITO-glass substrates were subject to wet (isopropyl alcohol) and dry (UV-ozone) cleaning processes before spin-coating. On top of the cleaned ITO-glass substrates, the hole-collecting buffer layer (HCBL) was spin-coated at 2500 rpm for 1 min using a conducting polymer solution with the major component poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) (PH500, HC Starck). The HCBL was further annealed at 230 °C for 15 min in order to circumvent any influence during subsequent annealing time experiments at 150 °C and 160 °C. Next, the P3HT:PC₆₁BM blend solutions were spun onto the HCBL at 1500 rpm for 30 s using a programmable spin-coater. After spin-coating, all samples were softbaked at 50 °C for 15 min. One set of the coated film samples was moved into a vacuum chamber system for depositing top electrodes. The aluminium (Al) electrode deposition was carried out at ∼1 × 10⁻⁶ Torr, leading to a 3 × 3 mm² active pixel area of devices. The completed devices were thermally annealed by varying annealing time at two different temperatures (150 °C and 160 °C) inside a nitrogen-filled glove box.

Synchrotron-radiation GIXD measurement

Another set of P3HT:PC₆₀BM blend film samples was taken into a GIXD measurement chamber system equipped in the Pohang Accelerator Laboratory (PAL, POSTECH, Republic of Korea). The wavelength and energy of the synchrotron X-ray beam was 0.138 nm and 8979 eV, respectively. The time-dependent measurement of 2D GIXD images was initialized by putting the film sample on the hot stage (in the vacuum chamber) where the temperature was controlled (150 °C or 160 °C) by employing a 2D X-ray image detector (Bicron scintillation counter, PI-SCX4300-165/2 2D CCD). A schematic diagram for GIXD measurement is shown in Fig. 1. The extraction of 1D profiles from 2D images was performed using home-made software.

Fig. 1 Schematic illustration for time-dependent 2D GIXD measurement of P3HT:PC₆₁BM blend films: I_IN, I_OOP, I_IP, θ_IN, 2θ, and α are incident X-ray intensity, diffracted X-ray intensity in out-of-plane (OOP) direction, diffracted X-ray intensity in the in-plane (IP) direction, the incident angle, diffraction angle, and the azimuthal angle, respectively.

Solar cell measurement and characterization

The current density–voltage (J--V) characteristics of P3HT:PC₆₀BM solar cells were measured using a specialized solar cell measurement system equipped with a solar simulator (Newport Oriel) and an electrometer (Keithley 2400). The intensity of simulated solar light was 100 mW cm⁻² after filtering through an air mass (AM) 1.5 G filter.

Results and discussion

As shown in Fig. 2, the 2D GIXD images gradually changed with the annealing time for both annealing temperatures (150 °C and 160 °C). At 150 °C the (100) Debye ring began to noticeably appear by only 2 min annealing and became more intense on further annealing, while the (200) spot appeared by 2 min annealing. This result roughly indicates that the P3HT crystallization proceeded continuously as the annealing time increased up to 120 min.

Fig. 2 2D GIXD images of P3HT:PC₆₁BM blend films as a function of annealing time at 150 °C: Note that the intense and broad rings above the (300) spot are diffractions mostly from the PEDOT:PSS-coated ITO-coated glass substrates in the presence of a small contribution of active layer components.¹⁹

However, the 2D GIXD images at 160 °C showed slightly different trends in terms of intensity (Fig. 3): The (100) Debye ring was already obviously shaped even at <1 min, while the (200) spot was also observed at <1 min. In particular, the (200) Debye ring was clearly measured for all annealing times except <1 min, indicating more pronounced high-order crystallization at higher temperature (i.e., 160 °C than 150 °C). The (300) spot began to appear at 2 min, which is almost identical to the result at 150 °C. Hence we consider that the higher order diffraction, which reflects more extended layer-by-layer ordering of P3HT crystallites in the P3HT:PC₆₁BM blend films, is made very initially (∼2 min) at the annealing temperature range of 140–160 °C which was found to be an optimum annealing temperature for P3HT:PC₆₁BM solar cells.^7–9,15,18

Fig. 3 2D GIXD images of P3HT:PC₆₁BM blend films as a function of annealing time at 160 °C: Note that the intense and broad rings above (300) spot are diffractions mostly from the PEDOT:PSS-coated ITO-coated glass substrates in the presence of a small contribution of active layer components.

For better comparison of GIXD results according to the annealing time at two different temperatures (150 °C and 160 °C), the 2D GIXD images were extracted into each 1D profile with respect to the azimuthal angle (α). As shown in Fig. 4, the (100) peak position was slightly shifted toward lower angles by only 2 min annealing irrespective of annealing temperature. However, further extension of annealing time did not change the (100) peak position. This result indicates that the P3HT crystallization (stacking) in the OOP direction was almost completed by short-time annealing at these temperatures, which is a similar result to the previous report where the annealing temperature was 140 °C.¹⁹ Hence it is considered that the quick rearrangement of P3HT stacking (crystallization) in the OOP direction is made actively at temperatures between 140 °C and 160 °C. We note that the (200) peak was also similarly shifted after 2 min annealing for both temperatures.

Fig. 4 The OOP direction 1D GIXD profiles of P3HT:PC₆₁BM blend films as a function of annealing time (t_A) at 150 °C (top panel) and 160 °C (bottom panel).

A similar initial shift of (100) peaks was observed for the 1D GIXD profiles in the IP direction (Fig. 5). Interestingly, however, the extent of (100) peak shift toward lower angles was more pronounced for annealing at 160 °C than 150 °C.

Fig. 5 The IP direction 1D GIXD profiles of P3HT:PC₆₁BM blend films as a function of annealing time (t_A) at 150 °C (top panel) and 160 °C (bottom panel).

This result implies that the stacking rearrangement (crystallization) of P3HT chains is more active at higher temperatures in the presence of PC₆₁BM molecules. Here it is noteworthy that much clearer (200) peaks were observed from the 1D profiles at higher temperature (160 °C).

To investigate the extent of P3HT stacking rearrangement (crystallization) with the annealing time, the (100) peak intensity was plotted as a function of annealing time. As shown in Fig. 6(a), the (100) peak in the OOP direction quickly increased as the annealing time increased up to 7 min at 150 °C. After this point, the intensity became marginally lower and then increased again leading to a maximum at 25 min. Then further annealing resulted in a gradually decreasing trend, which indicates a destruction of P3HT stacking (crystallization) in the OOP direction. In the case of 160 °C, the initial quick rise was also measured but the extent of the (100) intensity increase at 2 min was significantly different from that of 150 °C annealing [Fig. 6(b)]. This initial bigger jump in the intensity at 160 °C can be attributed to the more accelerated crystallization of P3HT chains at higher temperature.^21,22 Here it is noteworthy that the up–down–up fluctuations were observed for both temperatures, even though the strength of fluctuations was quite small compared to that of 140 °C annealing in our previous report.¹⁹ This small fluctuation in the (100) intensity can be ascribed to the influence of the higher annealing temperature. For both annealing temperatures, the (100) intensity exhibited a decreasing tendency after showing a maximum value at about 20–30 min. In particular, the decreasing tendency was more pronounced for 160 °C annealing than 150 °C. This result might reflect that the P3HT stacking in the OOP direction became weakened by long-time annealing at higher temperatures, which can be attributed to the randomization of P3HT chain orientations owing to the formation of larger PC₆₁BM clusters.²³ We note that the 140 °C annealing in the previous report¹⁹ exhibited only a marginal decay in the (100) intensity up to 120 min. This result provides us with a view of the annealing time–temperature correlation in P3HT:PC₆₁BM blend films, which might eventually affect the device performance.

Fig. 6 Change of (100) OOP peak intensity in the 1D GIXD profiles of P3HT:PC₆₁BM blend films as a function of annealing time (t_A) at 150 °C (a) and 160 °C (b).

As summarized in Table 1, the size of P3HT crystallites^19,24–26 in the P3HT:PC₆₁BM blend films was greatly changed with annealing time. The crystallite size in the OOP direction (L_OOP) was bigger for 160 °C annealing than 150 °C annealing (L_IP) at 5 min, whereas its trend was reversed for the IP direction. Interestingly, the 160 °C annealing resulted in almost similar crystallite sizes (the difference was less than 2.5 nm) in both directions, even though a larger crystallite size in the IP direction than the OOP direction was measured for both 150 °C and 140 °C annealing. Considering the general principle that larger populations of nuclei are generated at higher temperatures than lower temperatures, the smaller crystallite size in the IP direction at 160 °C could be attributed to the generation of more crystallites at this temperature than at either 140 °C or 150 °C.

Table 1 Summary of P3HT crystallite size changes in OOP (L_OOP) and IP (L_IP) directions for P3HT:PC₆₁BM blend films as a function of annealing time (t_A) at 150 °C and 160 °C

t A/min	L OOP/nm		L IP/nm
	150 °C	160 °C	150 °C	160 °C
<1	8.01	7.91	7.82	7.40
5	9.79	12.93	18.16	10.46
7	11.29	11.94	15.30	11.69
10	11.74	11.51	16.62	11.76
30	13.86	13.40	11.43	10.99
60	11.26	10.85	12.92	12.42
120	11.66	11.19	10.61	11.87

---

Based on the annealing time-dependent crystallization trend of P3HTs in the P3HT:PC₆₁BM blend films, which was sensitively dependent on the annealing temperature, we tried to investigate any device performance change according to the annealing time at two different temperatures. As shown in Fig. 7, the light J–V curves became better upon thermal annealing of devices. However, the 150 °C annealing exhibited a more pronounced change with annealing time than the 160 °C annealing. The detailed performance change is summarized in Fig. 8. The trend in the short-circuit current density (J_SC) was in good agreement with the (100) GIXD peak intensity trend in Fig. 6 when it comes to a more pronounced maximum J_SC (at around 30 min) and a quicker J_SC decaying trend at longer times for 150 °C annealing than 160 °C annealing. However, the open-circuit voltage (V_OC) was not changed with annealing time except for the non-annealed device in Fig. 8(a): We note that two different sets of non-annealed devices were used for each measurement, which inevitably led to the slightly different performance of the non-annealed devices in Fig. 8(a) and (b).

Fig. 7 Light J–V characteristics of P3HT:PC₆₁BM solar cells annealed at 150 °C (top panel) and 160 °C (bottom panel) by varying the annealing time (t_A): Only representative J–V curves related to major device performance changes are shown in order to avoid crowding figures.

Fig. 8 Change of device performances as a function of annealing time (t_A) at 150 °C (a) and 160 °C (b).

The fill factor (FF) trend also varied with annealing temperature, even though the initial quick increase was the same in spite of the different temperatures. This indicates that the nanostructure of P3HT:PC₆₁BM blend films inside the devices gradually changed with annealing time. However, the initial fluctuation (oscillation) in FF was less pronounced for the device annealed at the higher temperatures (150 °C and 160 °C) than at the lower temperature (140 °C in the previous work¹⁹). This less pronounced fluctuation can be attributed to the influence of the higher annealing temperature, which might accelerate the P3HT crystallization leading to a quicker arrival at an optimum stage. It is also worth noting that the power-conversion efficiency (PCE) mostly followed the FF trend without respect to the annealing temperature. As summarized in Table 2, the series resistance (R_S) trend was also in good agreement with the FF trend: The R_S value significantly decreased as the annealing time increased up to 30 min, whereas it exhibited up–down trends upon further annealing.

Table 2 Summary of device parameters for P3HT:PC₆₁BM solar cells as a function of annealing time (t_A) at two different annealing temperatures (T_A)

T A/°C	t A/min	V OC/V	J SC/mA cm^−2	FF (%)	PCE (%)	R S/kΩ cm^2
150	0	0.63	2.91	0.38	0.71	1.97
	7	0.59	10.53	0.52	3.23	0.11
	10	0.59	10.82	0.52	3.33	0.11
	30	0.59	11.57	0.53	3.68	0.11
	60	0.59	11.43	0.50	3.41	0.14
	120	0.59	10.05	0.52	3.11	0.12
160	0	0.60	3.21	0.42	0.81	1.04
	5	0.60	10.65	0.54	3.45	0.11
	7	0.60	10.32	0.49	3.08	0.18
	30	0.60	10.57	0.50	3.18	0.18
	60	0.59	10.34	0.48	2.98	0.19
	120	0.59	10.28	0.52	3.20	0.13

---

Conclusions

The crystallization of P3HT chains in P3HT:PC₆₁BM blend films was measured by varying the annealing time at two different annealing temperatures and employing a synchrotron-radiation GIXD technique. The 2D GIXD images measured at 150 °C disclosed that the higher order crystallization (stacking) of P3HT chains began to occur very quickly (after only 2 min). Considering the intense (100) Debye ring at 160 °C even after only 2 min, however, the P3HT crystallization had already proceeded considerably during this very short annealing time at this temperature. The 1D GIXD profiles showed that the (100) diffraction peak was shifted toward lower angle directions by initial short-time (2 min) annealing at higher temperatures (150 °C and 160 °C), as similarly observed for 140 °C annealing in the previous report. However, the (100) peak intensity exhibited quite a different trend with annealing temperature: The intensity fluctuation below 30 min was very small at 150 °C and 160 °C, compared to that at 140 °C (previous report), even though the intensity decay after 30 min became larger with annealing time at higher annealing temperatures. The trend of device performance with annealing time was in good accordance with that of the GIXD intensity change irrespective of annealing temperature. However, after an initial quick increase in device performance, no significant fluctuation was observed by further extension of annealing time at the present temperatures (150 °C and 160 °C), which is different from the trend reported in the previous work (140 °C annealing). This discrepancy with annealing temperature has been assigned to the temperature-dependent crystallization of P3HT in P3HT:PC₆₁BM blend films.

Acknowledgements

This work was financially supported by Korean government grants (Priority Research Center Program_2009-0093819, Pioneer Research Center Program_2010-0002231, NRF_20090072777, KETEP-2008-N-PV08-J-01-30202008, NRF_20100004164).

References

1. K. M. Coakley and M. D. McGehee, Chem. Mater., 2004, 16, 4533.
2. S. Günes, H. Neugebauer and N. S. Sariciftci, Chem. Rev., 2007, 107, 1324.
3. G. Dennler, M. C. Scharber and C. J. Brabec, Adv. Mater., 2009, 21, 1323.
4. L. Chen, Z. Hong, G. Li and Y. Yang, Adv. Mater., 2009, 21, 1434.
5. S. E. Shaheen, C. J. Brabec, N. S. Sariciftci, F. Padinger, T. Fromherz and J. C. Hummelen, Appl. Phys. Lett., 2001, 78, 841.
6. F. Padinger, R. S. Rittberger and N. S. Sariciftci, Adv. Funct. Mater., 2003, 13, 85.
7. Y. Kim, S. A. Choulis, J. Nelson, D. D. C. Bradley, S. Cook and J. R. Durrant, Appl. Phys. Lett., 2005, 86, 063502.
8. M. Reyes-Reyes, K. Kim and D. L. Carroll, Appl. Phys. Lett., 2005, 87, 083506.
9. W. Ma, C. Yang, X. Gong, K. Lee and A. J. Heeger, Adv. Funct. Mater., 2005, 15, 1617.
10. G. Li, V. Shrotriya, J. Huang, Y. Yao, T. Moriarty, K. Emery and Y. Yang, Nat. Mater., 2005, 4, 864.
11. J. H. Huang, F. C. Chien, P. Chen, K. C. Ho and C. W. Chu, Anal. Chem., 2010, 82, 1669.
12. G. D. Sharma, P. Suresh, S. S. Sharma, Y. K. Vijay and J. A. Mikroyannidis, ACS Appl. Mater. Interfaces, 2010, 2, 504.
13. S. van Bavel, E. Sourty, G. de With, S. Veenstra and J. Loos, J. Mater. Chem., 2009, 19, 5388.
14. J. H. Huang, C. Y. Yan, G. Z. Y. Ho, D. Kekuda, M. C. Wu, C. W. Chu and K. C. Ho, Org. Electron., 2009, 10, 27.
15. Y. Kim, S. Cook, S. M. Tuladhar, J. Nelson, J. R. Durrant, D. D. C. Bradley, M. Giles, I. McCulloch, C. S. Ha and M. Ree, Nat. Mater., 2006, 5, 197.
16. H. Y. Chen, J. Hou, S. Zhang, Y. Liang, G. Yang, Y. Yang, L. Yu, Y. Wu and G. Li, Nat. Photonics, 2009, 3, 649.
17. S. Park, A. Roy, S. Beaupre, S. Cho, N. Coates, J. Moon, D. Moses, M. Leclerc, K. Lee and A. J. Heeger, Nat. Photonics, 2009, 3, 297.
18. Y. Kim, J. Nelson, T. Zhang, S. Cook, J. R. Durrant, H. Kim, J. Park, M. Shin, S. Nam, M. Heeney, I. McCulloch, C. S. Ha and D. D. C. Bradley, ACS Nano, 2009, 3, 2557.
19. M. Shin, H. Kim, J. Park, S. Nam, K. Heo, M. Ree, C. S. Ha and Y. Kim, Adv. Funct. Mater., 2010, 20, 748.
20. H. Kim, M. Shin, J. Park and Y. Kim, ChemSusChem, 2010, 3, 476.
21. H. Yang, T. J. Shin, L. Yang, K. Cho, C. Y. Ryu and Z. Bao, Adv. Funct. Mater., 2005, 15, 671.
22. R. J. Kline, D. M. DeLongchamp, D. A. Fischer, E. K. Lin, L. J. Richter, M. L. Chabinyc, M. F. Toney, M. Heeney and I. McCulloch, Macromolecules, 2007, 40, 7960.
23. C. H. Woo, B. C. Thompson, B. J. Kim, M. F. Toney and J. M. J. Frechet, J. Am. Chem. Soc., 2008, 130, 16324.
24. T. Erb, U. Zhokhavets, G. Gobsch, S. Raleva, B. Stühn, P. Schilinsky, C. Waldauf and C. J. Brabec, Adv. Funct. Mater., 2005, 15, 1193.
25. M. Chiu, U. Jeng, C. Su, K. S. Liang and K. Wei, Adv. Mater., 2008, 20, 2573.
26. L. E. Alexander, X-ray Diffraction Methods in Polymer Science, John Wiley & Sons, New York, 1969.

---