**5.1 Power conversion efficiency and current-voltage dependence**

In order to clarify the effect of chemical additives on the photophysical properties and photovoltaic performance, regioregular P3HT and PCBM bulk-heterojunction solar cells were fabricated in four different ways:

(1) as produced films (untreated, no alkyl thiol);

18 Will-be-set-by-IN-TECH

Improved crystallization of P3HT and broader crystal size distribution at higher 1,8-octanedithiol concentrations was explained by solvent volume ratios. During the film fabrication, the main solvent evaporates faster than the additive solvent resulting in a sudden increase of the volume ratio of the additive solvent to the main solvent. Polymer molecules lower their internal energy by aggregating when the additive solvent volume ratio reaches a critical point. At higher additive concentrations, the time required to reach this point is reduced and aggregation is stronger. As a result, polymer molecules aggregate with larger average domain sizes due to the stronger driving force and broader size distributions arises

The morphological studies discussed above highlight the importance of phase separation between donor and acceptor, and reveal a schematic film structures for polymer-based bulk-heterojunction solar cells, as shown in Fig. 14..(Hoppe et al., 2006; Huang et al., 2010;

In the top Fig. 14 (a), the percolated pathways for electrons and holes is created allowing them to reach the respective electrodes. In Fig. 14 b the situation for an enclosed PCBM cluster is

The center Fig. 14 show that the lower surface energy of P3HT, relative to PCBM, provides the driving force for the interconcentration gradient observed in both the rapidly (a) and slowly (b) grown films. The film prepared through a rapidly grown process leads to an extremely homogeneous blends. A greater number of percolating pathways are formed in slow grown

Furthermore, the effect of annealing on the interface morphology of a mixed-layer device was modeled using a cellular model, as shown in Fig. 14 (bottom) for different temperatures. Annealing temperatures has been shown to crucially influence the morphology of the mixed-layer device, while the modeled morphology resemble experimentally measured

The photophysical effects of 1,8-octanedithiol (ODT) additives on PCPDTBT and C71-PCBM composites and device performance were studied using photo-induced absorption spectroscopy.(Hwang et al., 2008) Reduced carrier loss due to recombination was found in BHJ films processed using the additive. From photobleaching recovery measurements reduced carrier losses were demonstrated. However, it was concluded that the amount of the reduction is not sufficient to explain the observed increase in the power conversion efficiency (by a factor of 2). Carrier mobility measurements in Field Effect Transistor (FET) configuration demonstrated that the electron mobility increased in the PCPDTBT:C71-PCBM when ODT is used as an additive, resulting in enhanced connectivity of C71-PCBM networks.(Cho et al., 2008) This work also showed that if the ODT was not completely removed from the BHJ films by placing them in high vacuum (> 10−<sup>6</sup> torr) the hole mobility actually decreased, implying that residual ODT may act as a hole trap. It was concluded that the improved electron mobility was the primary cause of the improved power conversion efficiency, while the hole mobility

shown: here electrons and holes will recombine, since percolation is insufficient.

**5. Processing additive effect on solar cell performance**

was found to be relatively insensitive to the additive.

**4. Schematic structures of bulk-heterojunction film morphology**

due to the shorter aggregation time.

Peumans et al., 2003)

films.

devices.

(2) thermally annealed films (refereed to as treated in text, no alkyl thiol);

(3) as produced films with alkyl thiol (refereed to as treated in text, with alkyl thiol);

(4) thermally annealed films with alkyl thiol (refereed to as treated in text, with alkyl thiol).

The fabrication procedures were kept the same for all four types of cells. The details on device preparation can be found elsewhere.(Pivrikas et al., 2008)

Current-voltage (I-V) characteristics under illumination of devices are shown in Fig. 15. Untreated solar cells gave the worst performance with the least short circuit current and low fill factor. However, these cells demonstrate a relatively higher open circuit voltage, but, due to a low short circuit current and a low fill factor, their power conversion efficiency was low, around 1 %. The difference in photocurrents between annealed cells and these with alkyl thiol

Fig. 14. Schematic structures of the film nanomorphology of bulk-heterojunction blends - all emphasizing the importance of the interpenetrating network in polymer-based solar cells. Top figures: (a) chlorobenzene and (b) toluene cast MDMO-PPV and PCBM blend layers. Center figures: vertical phase morphology of (a) rapidly and (b) slowly grown P3HT and PCBM blends. Bottom figures: the simulated effects of annealing on the interface morphology of a mixed-layer photovoltaic cell. The interface between donor and acceptor is shown as a green surface. Donor is shown in red and acceptor is transparent. Top figures reprinted with permission from (Hoppe et al., 2006), copyright 2006, with permission from Elsevier. Middle figures reprinted with permission from (Huang et al., 2010), copyright 2010 American Chemical Society. Bottom figures adapted by permission from Macmillan Publishers Ltd: (Peumans et al., 2003), copyright 2003.

Fig. 16. Changes in light absorption (a) and photoluminescence (PL) (b) and External Quantum Efficiency (EQE) (c) shown at various amounts of processing additive (OT is 1,8-octanedithiol) used during film preparation. Changes in light absorption (d) and incident photon to current efficiency (IPCE) in (e) measured in pristine and treated (annealed films and films fabricated with processing additive) films. Strong red-shift in absorption,

from (Pivrikas et al., 2008). Copyright 2008, with permission from Elsevier.

**5.3 Charge transport**

appearance of absorption peaks, higher IPCE values in treated films or films with processing additive well agrees with improved OPV performance. Thermal annealing of films fabricated with processing additive results in no change in OPV performance. Figures on the left reprinted with permission from (Chen, Yang, Yang, Sista, Zadoyan, Li & Yang, 2009). Copyright 2009 American Chemical Society. Figures on the right reprinted with permission

Relation Between Nanomorphology and Performance of Polymer-Based Solar Cells 141

A strong improvement in IPCE was observed in treated solar cells. The IPCE dependence approximately follows the light absorption curve, as the same characteristic absorption peaks are reproduced in the optical absorption spectra (Fig. 16). From the IPCE studies it was concluded that the improvement in the performance of solar cells is not only due to the increased optical absorption, but also due to improved transport (higher carrier mobility) and/or reduced recombination losses (eg. due to longer charge carrier lifetime), which again confirms the benefits of improved interpenetrating network between donor and acceptor.

Since it was found from ICPE studies that the film morphology not only improves the light absorption, but also results in better charge transport, it is important to quantify this improvement. In order to understand the difference in charge transport properties in treated

Fig. 15. Current-voltage characteristics demonstrating significant performance improvement under illumination (1000 W/m2, 1.5 AM) for P3HT/PCBM bulk-heterojunction solar cells prepared in different ways: as produced (thin line), annealed (thick dashed line), thiol added (thick line), thiol added and annealed (thick dash dot line). Reprinted with permission from (Pivrikas et al., 2008). Copyright 2008, with permission from Elsevier.

is small, except that treated cells have lower fill factors and therefore slightly lower efficiency as compared to those with alkyl thiol additive, Fig. 16.

#### **5.2 Light absorption and external quantum efficiency**

In order to clarify the factors determining OPV device efficiency, the incident photon to current efficiency (IPCE), alternatively called External Quantum Efficiency (EQE) is measured, since it provides information on light absorption spectra, charge transport and recombination losses. The effect of thermal treatment versus processing addictive, as well as the effect of additive concentration, was studied and shown in Fig. 16. In Fig. 16 (a) and (d) the light absorption and Beer-Lambert absorption coefficient are shown as a function of wavelength. In agreement with previous observations, an increase in optical absorption is seen for treated cells. The red-shift of the absorption and characteristic vibronic shoulders are clearly pronounced in treated cells (at around 517 nm, 556 nm and 603 nm) both arising from strong interchain interactions within high degree of crystallinity in P3HT. In solution, no peak shift was observed, suggesting that the influence of the additive on P3HT happens during the solvent drying (or spin coating) process and not in the solution state. The increase in optical absorption at higher additive concentrations demonstrates that more energy can be harvested in solar cells, therefore, these cells have better photovoltaic performance due to a larger amount of photons being absorbed in the film.

While PCBM is known to quench the PL of P3HT effectively in the well mixed blends.(Chen, Yang, Yang, Sista, Zadoyan, Li & Yang, 2009) The photoluminescence was shown to increase with increasing amount of 1,8-octanedithiol (Fig. 16 (b)), suggesting that the phase separation between the P3HT and PCBM is increasing since the exciton diffusion distance is on the same order of magnitude.(Xu & Holdcroft, 1993; Zhokhavets et al., 2006)

Fig. 16. Changes in light absorption (a) and photoluminescence (PL) (b) and External Quantum Efficiency (EQE) (c) shown at various amounts of processing additive (OT is 1,8-octanedithiol) used during film preparation. Changes in light absorption (d) and incident photon to current efficiency (IPCE) in (e) measured in pristine and treated (annealed films and films fabricated with processing additive) films. Strong red-shift in absorption, appearance of absorption peaks, higher IPCE values in treated films or films with processing additive well agrees with improved OPV performance. Thermal annealing of films fabricated with processing additive results in no change in OPV performance. Figures on the left reprinted with permission from (Chen, Yang, Yang, Sista, Zadoyan, Li & Yang, 2009). Copyright 2009 American Chemical Society. Figures on the right reprinted with permission from (Pivrikas et al., 2008). Copyright 2008, with permission from Elsevier.

A strong improvement in IPCE was observed in treated solar cells. The IPCE dependence approximately follows the light absorption curve, as the same characteristic absorption peaks are reproduced in the optical absorption spectra (Fig. 16). From the IPCE studies it was concluded that the improvement in the performance of solar cells is not only due to the increased optical absorption, but also due to improved transport (higher carrier mobility) and/or reduced recombination losses (eg. due to longer charge carrier lifetime), which again confirms the benefits of improved interpenetrating network between donor and acceptor.

#### **5.3 Charge transport**

20 Will-be-set-by-IN-TECH

Fig. 15. Current-voltage characteristics demonstrating significant performance improvement under illumination (1000 W/m2, 1.5 AM) for P3HT/PCBM bulk-heterojunction solar cells prepared in different ways: as produced (thin line), annealed (thick dashed line), thiol added (thick line), thiol added and annealed (thick dash dot line). Reprinted with permission from

is small, except that treated cells have lower fill factors and therefore slightly lower efficiency

In order to clarify the factors determining OPV device efficiency, the incident photon to current efficiency (IPCE), alternatively called External Quantum Efficiency (EQE) is measured, since it provides information on light absorption spectra, charge transport and recombination losses. The effect of thermal treatment versus processing addictive, as well as the effect of additive concentration, was studied and shown in Fig. 16. In Fig. 16 (a) and (d) the light absorption and Beer-Lambert absorption coefficient are shown as a function of wavelength. In agreement with previous observations, an increase in optical absorption is seen for treated cells. The red-shift of the absorption and characteristic vibronic shoulders are clearly pronounced in treated cells (at around 517 nm, 556 nm and 603 nm) both arising from strong interchain interactions within high degree of crystallinity in P3HT. In solution, no peak shift was observed, suggesting that the influence of the additive on P3HT happens during the solvent drying (or spin coating) process and not in the solution state. The increase in optical absorption at higher additive concentrations demonstrates that more energy can be harvested in solar cells, therefore, these cells have better photovoltaic performance due to a larger amount of photons being absorbed

While PCBM is known to quench the PL of P3HT effectively in the well mixed blends.(Chen, Yang, Yang, Sista, Zadoyan, Li & Yang, 2009) The photoluminescence was shown to increase with increasing amount of 1,8-octanedithiol (Fig. 16 (b)), suggesting that the phase separation between the P3HT and PCBM is increasing since the exciton diffusion distance is on the same

(Pivrikas et al., 2008). Copyright 2008, with permission from Elsevier.

order of magnitude.(Xu & Holdcroft, 1993; Zhokhavets et al., 2006)

as compared to those with alkyl thiol additive, Fig. 16.

**5.2 Light absorption and external quantum efficiency**

in the film.

Since it was found from ICPE studies that the film morphology not only improves the light absorption, but also results in better charge transport, it is important to quantify this improvement. In order to understand the difference in charge transport properties in treated

**6. Conclusions**

is required.

preparation.

**8. References**

**7. Acknowledgements**

Wiley & Sons Inc.

*Physics Letters* 93: 013306.

*solar cells* 83(2-3): 273–292.

*Materials* 11(1): 15–26.

The film nanomorphology of bulk heterojunction solar cells determines the power conversion efficiency through photophysical properties such as light absorption, exciton dissociation, charge transport and recombination. The nano-morphology can be controlled by a variety of different methods. Thermal annealing of fabricated solar cells can be successfully substituted with slow drying of the solvent or chemical additives. These methods induce the phase separation between the donor and acceptor in the bulk-heterojunction, which results in red-shifted light absorption, improved exciton dissociation, faster charge carrier transport, and reduced recombination. Segregated donor-enriched and/or acceptor-enriched phases can be formed resulting in an interpenetrating bicontinuous network with the domain sizes comparable to the exciton diffusion length. Interconnected pathways for electromn and hole transport to the electrodes are required. This structure is essential for the photovoltaic performance of polymer-based solar cells. Therefore, reproducible, low cost nano-structure control is crucially important for fabrication of high efficiency OPV suitable for commercialization. In order to be able to control and predict the film nano-morphology of novel materials, an understanding of the material parameters governing the phase separation

Relation Between Nanomorphology and Performance of Polymer-Based Solar Cells 143

The author would like to Dr. Paul Schwenn for helpful discussions during manuscript

Alargova, R., Deguchi, S. & Tsujii, K. (2001). Stable colloidal dispersions of fullerenes in polar organic solvents, *Journal of the American Chemical Society* 123(43): 10460–10467. Baranovski, S. (2006). *Charge transport in disordered solids with applications in electronics*, John

Bassler, H. (1993). Charge transport in disordered organic photoconductors a monte carlo

Bechara, R., Leclerc, N., Lévêque, P., Richard, F., Heiser, T. & Hadziioannou, G. (2008).

Brabec, C. (2004). Organic photovoltaics: technology and market, *Solar energy materials and*

Brabec, C., Sariciftci, N., Hummelen, J. et al. (2001). Plastic solar cells, *Advanced Functional*

Campbell, A., Hodgkiss, J., Westenhoff, S., Howard, I., Marsh, R., McNeill, C., Friend, R. &

Chen, F., Tseng, H. & Ko, C. (2008). Solvent mixtures for improving device efficiency of

Chen, F., Xu, Q. & Yang, Y. (2004). Enhanced efficiency of plastic photovoltaic devices by blending with ionic solid electrolytes, *Applied physics letters* 84: 3181.

performance polymer photovoltaics, *Nano letters* 8(11): 3942–3947.

polymer photovoltaic devices, *Applied Physics Letters* 92: 103316.

Efficiency enhancement of polymer photovoltaic devices using thieno-thiophene based copolymers as nucleating agents for polythiophene crystallization, *Applied*

Greenham, N. (2008). Low-temperature control of nanoscale morphology for high

simulation study, *physica status solidi (b)* 175(1): 15–56.

and untreated cells, dark IV curves were recorded for all 4 types of treated cells shown in Fig. 17.(Pivrikas et al., 2008)

Fig. 17. The improvement in charge carrier mobility in treated (annealed films and films fabricated with processing additive) compared to pristine films demonstrated by two methods: dark current-voltage injection and CELIV. (a) log-lin plot showing the rectification ratio in forward and reverse bias and insignificant differences in leakage current in reverse bias. (b) log-log plot in forward bias showing much higher injection current levels in treated blends. (c) faster carrier extraction in treated films compared to pristine directly measured by CELIV current transients. Improvement in the carrier mobility can be seen from the shift in the position of extraction maximum, while experimental conditions (film thicknesses and applied voltages ) were kept similar. Thermal annealing of films fabricated with processing additive results in no change in performance. Reprinted with permission from (Pivrikas et al., 2008). Copyright 2008, with permission from Elsevier.

The dark current in the region of negative applied voltage (the reverse bias, positive voltage on Al, negative on ITO), is similar in all cells, showing that current injection is contact limited. A significant rectification ratio is observed for all types of studied cells. The dark leakage current in reverse bias is rather high, but similar for all cells.

Due to the different nanomorphologies of the interpenetrating network, the dark conductivity is expected to increase in the cells with higher conversion efficiency, because of improved conductivity of the films (assuming the injection is not limited by the contact). The dark injection current in forward bias is observed to be significantly higher in treated cells. In Fig. 17 (b) the dark injection current in forward bias is plotted in log-log scale for all devices. Faster charge carrier mobilities in all cells were estimated from these dependences using the Mott-Gurney Law. As can be directly seen from the magnitude of injection current, the highest mobility was observed in the films with chemical additives, confirming the beneficial effect of chemical additives for charge transport in bulk-heterojunction solar cells. From CELIV measurements, shown in Fig. 17 (c) it was demonstrated that charge carrier mobility is mainly reponsible for improvements in OPV performance.

However, the charge carrier recombination processes in operating devices has yet to be clarified. It was shown that the typically expected Langevin bimolecular charge carrier recombination can be avoided in highly efficiency P3HT and PCBM blends.(Pivrikas et al., 2005) Non-Langevin carrier recombination was shown to be crucially important in low mobility organic photovoltaic devices, since the requirement for the slower carrier mobility can be reduced without recombination losses. This implies that close to unity Internal quantum efficiency can be reached in low bandgap organic materials with very low carrier mobility if reduced bimolecular recombination (non-Langevin) is present in the device.
