**2.2 Solar cell fabrication techniques**

2 Will-be-set-by-IN-TECH

nanomorphology of polymeric solar cells plays a crucial role for the performance of the

Historically, thermal annealing of the film has been used to induce the phase separation between donor and acceptor in bulk-heterojunction blends.(Padinger et al., 2003) However, thermal treatment creates an additional fabrication step in the whole device fabrication process. Later, various methods have been tested and employed to control the nanomorphology of the blends, namely use of solvents with different boiling points (choice of solvent), reduction of drying speed (rate of drying and vapor annealing), changing the solubility of materials, melting of bilayers and the use of processing additives.(Pivrikas et al., 2010b) The later method has received great academic interest as it removes the need for post-production treatment while at the same time allowing fine control of the

In this work the factors limiting the power conversion efficiency of excitonic polymer-based bulk-heterojunction solar cells are discussed. Various methods allowing the film nanomorphology to be controlled are reviewed. The use of processing additives to control the phase separation for the formation of an interpenetrating network and how this impacts

Polymer-based bulk-heterojunction solar cells (BHSC) have already shown certified efficiencies above 8 % demonstrating ability to compete with inorganic solar cell systems (eg. amorphous silicon cells fabricated on flexible substrates). Efficiencies exceeding 10% for solution processed solar cells are expected to be achieved soon.(Nayak et al., 2011) The power conversion efficiency of BHSCs is determined by the photophysical processes under operational conditions. A fundamental understanding of the relation between light absorption charge separation, charge transport, recombination, and film nanostructure as well as between the various thin film fabrication and processing parameters (such as solvent composition, solution concentration, deposition atmosphere and process temperature) is needed for further improvements. These important parameters can be controlled to some extent by adjusting the required film composition or device structure.(Gunes et al., 2007)

The typical device structure of most organic optoelectronic devices, including organic light-emitting diodes and solar cells, is shown in Fig. 1. The front electrode is based on a transparent conducting oxide, such as indium tin oxide (ITO), that serves as the high-work-function, positive electrode.(Brabec et al., 2001) To further improve the quality of the ITO electrode and aid hole (positive charge carrier) extraction from the active film, poly(3,4-ethylenedioxythiophene)-poly(styrene sulfonate) (PEDOT-PSS) layer (tens of nanometers thick) is coated on top, forming a smooth surface which is essential in thin film devices. The photoactive film, the donor and acceptor bulk-heterojunction blend is deposited on top of the PEDOT-PSS layer. The whole device is finished by thermally evaporating the back contact (negative electrode) under high vacuum. To achieve the built-in electric field needed for most devices to operate, the back electrode must be made from a low-work-function metal that serves as the negative electrode. In the operation of a typical polymer-fullerene bulk-heterojunction solar cell, electrons generated in the active layer are collected by the back electrode (anode), and holes are collected at the opposite electrode

nanomorphology in various donor-acceptor blends.(Lee et al., 2008)

the power conversion efficiency is described.

**2. Excitonic polymer-based solar cells**

**2.1 Solar cell device structure**

(cathode).

devices.

While academic research is highly concentrated on improving the power conversion efficiency, there are other important aspects needed for commercial success, such as cell stability, degradation, low manufacturing costs with rapid large scale production. This has been summarized by the Venn diagram as the unification of challenges when trying to combine power conversion efficiency, processability and stability into final devices.(Jorgensen et al., 2008)

Solution processing is attractive for fabricating organic optoelectronic devices mainly due to its simplicity and applicability for large scale and low-cost production. Thin films can be formed in various ways: a) printing techniques including screen printing, pad printing, gravure printing, flexographic printing and offset printing; b) coating techniques including pin coating, doctor blading, casting, painting, spray coating, slot-die coating, curtain coating, slide coating and knife-over-edge coating. The only technique that in both categories is inkjet printing.

Spin coating has been the most common technique for polymeric solar cell fabrication with numerous reviews and fundamental studies available.(Norrman et al., 2005) This technique, widely used in the microelectronics industry to deposit photoresist on silicon wafers, allows for the reproducible formation of highly homogeneous films over large areas. A typical spin coating process involves application of a solution (with the organic semiconductors dissolved in a solvent) to a substrate which is then either accelerated to the required angular velocity or is already spinning at it, Fig. 2.(Krebs, 2009) A large portion of the solution is wasted leaving a thin film on the substrate. Film thickness, morphology and surface topography strongly depend on the rotational speed, viscosity, volatility, diffusivity, molecular weight and concentration of the solutes and solvents used.(Cohen & Gutoff, 1992)

### **2.3 Current-Voltage dependence of solar cells**

The most important figure of merit describing the performance of a solar cell is the power conversion efficiency, which is determined from the current voltage characteristics of the solar cells under operational conditions. Typical current-voltage characteristics of solar cells under illumination is shown in Fig. 3.(Deibel & Dyakonov, 2010)

The accurate measurement of the PCE according to international standards has been described in the literature,(Shrotriya et al., 2006) and is eesential for reproducibility and comparison of results between different laboratories. The Shockley diode equation describes the

Fig. 1. Schematic sandwich-type structure of organic solar cells showing an organic semiconductor active film between two metal electrodes with different work functions (typically ITO/PEDOT-PSS as positive and Ca/Al as negative contacts). Reprinted with permission from (Shaheen, 2007). Copyright 2007 Society of Photo-Optical Instrumentation Engineers.

Fig. 3. Typical schematic current-voltage (I-V) dependence of the organic solar cell in the dark and under illumination. Reprinted with permission from (Deibel & Dyakonov, 2010).

Relation Between Nanomorphology and Performance of Polymer-Based Solar Cells 125

PCE is defined as the ratio of the electrical power produced by a solar cell to the optical power

*PCE* <sup>=</sup> *jSCVOCFF Plight*

where *jSC* is the short circuit current density, *VOC* is the open circuit voltage, *FF* = *VMPjmp*/*VOCjSC* is the fill factor representing the maximum area in the fourth quadrant of the I-V characteristics of solar cells, *VMP* and jmp are the voltage and current, respectively, at the point of maximum power, and *Plight* is the power of incoming light under Standard Test Conditions (1000 W/m2, AM 1.5 (Air Mass) solar reference spectrum, temperature during measurements 25 C).[15] To maximize the PCE values, all these parameters have to be maximized. The open circuit voltage of BHSC is determined by the energy of the quasi-Fermi levels of both semiconductors (donor and acceptor) as well as the Fermi levels of the electrodes.(Scharber et al., 2006) The short circuit current *jSC* depends on electrical carrier drift (induced by electric field) and diffusion (induced by concentration gradient). Apart from the absorption on the film, charge carrier concentration and mobility are the main factors influencing the photocurrent regardless of the transport mechanism (drift or

(1)

Copyright 2010, with permission from Institute of Physics.

of incident light (*Plight*):(Luque & Hegedus, 2003)

**2.4 Power conversion efficiency**

diffusion).(Nelson, 2003)

Fig. 2. Spin coating of organic solar cells from solution. Reprinted with permission from (Krebs, 2009). Copyright 2009, with permission from Elsevier.

current-voltage dependence of an ideal diode.(Shockley & Queisser, 1961) In the dark under forward bias the injection current increases exponentially with applied bias, whereas under reverse bias, current saturates at low applied voltages due to blocking contacts. This leads to a rectifying behaviour as can be clearly seen in the log-lin plot in Fig. 3. A description of the non-ideal device (typical organic solar cells) requires addition of series (Rs) and paralell (Rp) resistances. Rs is connected in series with the ideal diode and it describes the contact resistances such as injection barriers and sheet resistances. Rp arises due to the influence of local shunts between the two electrodes, i.e. additional current paths circumventing the diode. Typically in organic solar cells, a strong photocurrent dependence of applied electric field is observed manifesting as a non-saturated current at -1V reverse bias (Fig. 3 ). The field-dependent photocurrent arises due to:

a) field dependent mobile charge carrier generation, since an exciton has to dissociate into mobile carriers;(Oesterbacka et al., 2010)

b) charge carrier collection due to Hecht's law, if the extracted charge saturates at electric fields wheere the film thickness is larger than the carrier drift distance.(Hecht, 1932)

c) electric field dependent carrier mobility.(Pivrikas, Ullah, Sitter & Sariciftci, 2011)

In addition to electric field dependent mobility, the charge transport in disordered organic solar cells is also carrier concentration dependent. This effect arises due to the hopping nature of charge transport, where at higher carrier concentrations the carrier hopping probability between localized states increases (loosely speaking due to a higher density of localized states resulting in better electron wavefunction overlap) and therefore the carrier mobility increases.(PIVRIKAS, ULLAH, SINGH, SIMBRUNNER, MATT, SITTER & SARICIFTCI, 2011)

Fig. 3. Typical schematic current-voltage (I-V) dependence of the organic solar cell in the dark and under illumination. Reprinted with permission from (Deibel & Dyakonov, 2010). Copyright 2010, with permission from Institute of Physics.

#### **2.4 Power conversion efficiency**

4 Will-be-set-by-IN-TECH

Fig. 2. Spin coating of organic solar cells from solution. Reprinted with permission from

current-voltage dependence of an ideal diode.(Shockley & Queisser, 1961) In the dark under forward bias the injection current increases exponentially with applied bias, whereas under reverse bias, current saturates at low applied voltages due to blocking contacts. This leads to a rectifying behaviour as can be clearly seen in the log-lin plot in Fig. 3. A description of the non-ideal device (typical organic solar cells) requires addition of series (Rs) and paralell (Rp) resistances. Rs is connected in series with the ideal diode and it describes the contact resistances such as injection barriers and sheet resistances. Rp arises due to the influence of local shunts between the two electrodes, i.e. additional current paths circumventing the diode. Typically in organic solar cells, a strong photocurrent dependence of applied electric field is observed manifesting as a non-saturated current at -1V reverse bias (Fig. 3 ). The

a) field dependent mobile charge carrier generation, since an exciton has to dissociate into

b) charge carrier collection due to Hecht's law, if the extracted charge saturates at electric fields

In addition to electric field dependent mobility, the charge transport in disordered organic solar cells is also carrier concentration dependent. This effect arises due to the hopping nature of charge transport, where at higher carrier concentrations the carrier hopping probability between localized states increases (loosely speaking due to a higher density of localized states resulting in better electron wavefunction overlap) and therefore the carrier mobility increases.(PIVRIKAS, ULLAH, SINGH, SIMBRUNNER, MATT, SITTER & SARICIFTCI, 2011)

wheere the film thickness is larger than the carrier drift distance.(Hecht, 1932) c) electric field dependent carrier mobility.(Pivrikas, Ullah, Sitter & Sariciftci, 2011)

(Krebs, 2009). Copyright 2009, with permission from Elsevier.

field-dependent photocurrent arises due to:

mobile carriers;(Oesterbacka et al., 2010)

PCE is defined as the ratio of the electrical power produced by a solar cell to the optical power of incident light (*Plight*):(Luque & Hegedus, 2003)

$$PCE = \frac{j\_{\text{SC}}V\_{\text{OC}}FF}{P\_{\text{light}}} \tag{1}$$

where *jSC* is the short circuit current density, *VOC* is the open circuit voltage, *FF* = *VMPjmp*/*VOCjSC* is the fill factor representing the maximum area in the fourth quadrant of the I-V characteristics of solar cells, *VMP* and jmp are the voltage and current, respectively, at the point of maximum power, and *Plight* is the power of incoming light under Standard Test Conditions (1000 W/m2, AM 1.5 (Air Mass) solar reference spectrum, temperature during measurements 25 C).[15] To maximize the PCE values, all these parameters have to be maximized. The open circuit voltage of BHSC is determined by the energy of the quasi-Fermi levels of both semiconductors (donor and acceptor) as well as the Fermi levels of the electrodes.(Scharber et al., 2006) The short circuit current *jSC* depends on electrical carrier drift (induced by electric field) and diffusion (induced by concentration gradient). Apart from the absorption on the film, charge carrier concentration and mobility are the main factors influencing the photocurrent regardless of the transport mechanism (drift or diffusion).(Nelson, 2003)

materials (relative static permittivity is around 3) the primary photoexcitation is an exciton, which does not create the photocurrent.(Luque & Hegedus, 2003) The exciton diffusion length describes how far an exciton can diffuse within its lifetime. The concept of heterojunction between two organic semiconductors (donor and acceptor) is used to split an exciton into mobile charge carriers. Efficient exciton dissociation (charge transfer) takes place at the interface between donor and acceptor if a suitable offset in energy level exists, as shown in Fig. 4 b).(Pivrikas et al., 2010a) The excited state Charge Transfer (CT) complex (sometimes called exciplex) might be formed after the dissociation of an exciton meaning that positive and negative charge carriers might remain bounded by the Coulomb attraction at the donor

Relation Between Nanomorphology and Performance of Polymer-Based Solar Cells 127

The charge carrier transport (collection) to the electrodes is the third important limiting processes. As shown in Fig. 4 c), mobile electrons and holes must be transported to the opposite electrodes. The driving force can be either diffusion, related to the carrier concentration gradients, and/or drift due to a built-in electric field. An important aspects of charge transport are the charge extraction at the semiconductor-metal interface. The energy level alignment between the metal and semiconductor, free charge carrier concentration in the film (doping level) as well as the trapping level concetration, carrier capture and release times from capture centers determine the interfacial properties of the device. Non-blocking contact without energetical barrier for charge carrier extraction is required to be present at the interface.(Baranovski, 2006) The disordered nature of solution processed films of organic semiconductors results in low charge carrier mobilities (tzpically 10−<sup>3</sup> - 10−<sup>7</sup> cm2V−1s−<sup>1</sup> in *π*-conjugated polymers). The mobility of the slower charge carrier limits the photocurrent, and therefore the efficiency of the solar cell due to accumulation of charge carriers. Since the photocurrent under operation conditions typically approach space charge limited current, second order recombination processes become dominant due to high charge carrier concentration, and the carrier lifetime becomes shorter than the transit time.(Pivrikas

The light absorption coefficient, *α*, in a disordered organic film is usually high, on the order of 105 cm−1. This allows thin films, on the order of hundreds of nanometers, to be used in solar cells. However, the exciton diffusion length in most organic materials is of the order of 10 nm. If the exciton is to diffuse to the interface between the two materials (donor-acceptor) in order to separate into mobile charge carriers, these two materials must be blended on this length scale. Furthermore, the donor-acceptor phases must for bi-continuous network with percolating pathways for electron and hole transport to the elctrodes. This is the operating principle of the BHSC shown in Fig. 5.(Sariciftci, 2006) The film nano-morphology is crucially important for the efficiency of solar cells.(Ma et al., 2005) The nanoscale phase-separation phase separation between donor and acceptor in BHSC plays an important role relating the device properties and performance to the solar cell fabrication methods. Typical donor is poly(3-hexylthiophene-2,5-diyl) (P3HT) and acceptor is [6,6]-phenyl-C61-butyric acid methyl

The formation and the size of nanoscale domains of donor and acceptor phases are strongly dependent on the film fabrication techniques and conditions. Beyond the selection of suitable materials there are several parameters that must be carefully controlled when fabricating

acceptor interface, which would not contribute to photocurrent.

et al., 2010a)

ester (PCBM)

**2.6 Bulk-heterojunction solar cells**

**3. Methods to control the morphology of BHSC**

Fig. 4. Factors limiting power conversion efficiency of excitonic solar cells. Reprinted with permission from (Pivrikas, 2010). Copyright 2010 IEEE.

The PCE of organic solar cells is influenced by many different photophysical processes and parameters. Simplified energy-level diagrams of organic solar cells utilizing donor and acceptor materials is shown in Fig. 2.(Pivrikas et al., 2010a)

### **2.5 Power conversion efficiency limiting mechanisms in excitonic solar cells**

The first factor limiting the PCE is the absorption of light in the film. Ideally, as much as possible of the incident solar irradiance should be absorbed.(Pivrikas et al., 2010a) The Beer-Lambert law determines the light absorption profile in homogeneously distributed and scatter-free medium. Optical interference effects can also influence the light absorption profile in thin multilayer films.(Dennler et al., 2009) As can be seen in Fig. 4 some part of absorbed light energy is lost due to a thermalization process - charge relaxtion within the Density of States (DOS) to the lower energy levels to form an occupational-DOS within localized DOS.(Bassler, 1993; Juska et al., 2003; Osterbacka et al., 2003)

The second efficiency limiting process is exciton dissociation into mobile charge carriers. Due to low dielectric constants and consequently weak Coulombic field screening in the organic materials (relative static permittivity is around 3) the primary photoexcitation is an exciton, which does not create the photocurrent.(Luque & Hegedus, 2003) The exciton diffusion length describes how far an exciton can diffuse within its lifetime. The concept of heterojunction between two organic semiconductors (donor and acceptor) is used to split an exciton into mobile charge carriers. Efficient exciton dissociation (charge transfer) takes place at the interface between donor and acceptor if a suitable offset in energy level exists, as shown in Fig. 4 b).(Pivrikas et al., 2010a) The excited state Charge Transfer (CT) complex (sometimes called exciplex) might be formed after the dissociation of an exciton meaning that positive and negative charge carriers might remain bounded by the Coulomb attraction at the donor acceptor interface, which would not contribute to photocurrent.

The charge carrier transport (collection) to the electrodes is the third important limiting processes. As shown in Fig. 4 c), mobile electrons and holes must be transported to the opposite electrodes. The driving force can be either diffusion, related to the carrier concentration gradients, and/or drift due to a built-in electric field. An important aspects of charge transport are the charge extraction at the semiconductor-metal interface. The energy level alignment between the metal and semiconductor, free charge carrier concentration in the film (doping level) as well as the trapping level concetration, carrier capture and release times from capture centers determine the interfacial properties of the device. Non-blocking contact without energetical barrier for charge carrier extraction is required to be present at the interface.(Baranovski, 2006) The disordered nature of solution processed films of organic semiconductors results in low charge carrier mobilities (tzpically 10−<sup>3</sup> - 10−<sup>7</sup> cm2V−1s−<sup>1</sup> in *π*-conjugated polymers). The mobility of the slower charge carrier limits the photocurrent, and therefore the efficiency of the solar cell due to accumulation of charge carriers. Since the photocurrent under operation conditions typically approach space charge limited current, second order recombination processes become dominant due to high charge carrier concentration, and the carrier lifetime becomes shorter than the transit time.(Pivrikas et al., 2010a)

### **2.6 Bulk-heterojunction solar cells**

6 Will-be-set-by-IN-TECH

Fig. 4. Factors limiting power conversion efficiency of excitonic solar cells. Reprinted with

The PCE of organic solar cells is influenced by many different photophysical processes and parameters. Simplified energy-level diagrams of organic solar cells utilizing donor and

The first factor limiting the PCE is the absorption of light in the film. Ideally, as much as possible of the incident solar irradiance should be absorbed.(Pivrikas et al., 2010a) The Beer-Lambert law determines the light absorption profile in homogeneously distributed and scatter-free medium. Optical interference effects can also influence the light absorption profile in thin multilayer films.(Dennler et al., 2009) As can be seen in Fig. 4 some part of absorbed light energy is lost due to a thermalization process - charge relaxtion within the Density of States (DOS) to the lower energy levels to form an occupational-DOS within localized

The second efficiency limiting process is exciton dissociation into mobile charge carriers. Due to low dielectric constants and consequently weak Coulombic field screening in the organic

**2.5 Power conversion efficiency limiting mechanisms in excitonic solar cells**

permission from (Pivrikas, 2010). Copyright 2010 IEEE.

acceptor materials is shown in Fig. 2.(Pivrikas et al., 2010a)

DOS.(Bassler, 1993; Juska et al., 2003; Osterbacka et al., 2003)

The light absorption coefficient, *α*, in a disordered organic film is usually high, on the order of 105 cm−1. This allows thin films, on the order of hundreds of nanometers, to be used in solar cells. However, the exciton diffusion length in most organic materials is of the order of 10 nm. If the exciton is to diffuse to the interface between the two materials (donor-acceptor) in order to separate into mobile charge carriers, these two materials must be blended on this length scale. Furthermore, the donor-acceptor phases must for bi-continuous network with percolating pathways for electron and hole transport to the elctrodes. This is the operating principle of the BHSC shown in Fig. 5.(Sariciftci, 2006) The film nano-morphology is crucially important for the efficiency of solar cells.(Ma et al., 2005) The nanoscale phase-separation phase separation between donor and acceptor in BHSC plays an important role relating the device properties and performance to the solar cell fabrication methods. Typical donor is poly(3-hexylthiophene-2,5-diyl) (P3HT) and acceptor is [6,6]-phenyl-C61-butyric acid methyl ester (PCBM)

### **3. Methods to control the morphology of BHSC**

The formation and the size of nanoscale domains of donor and acceptor phases are strongly dependent on the film fabrication techniques and conditions. Beyond the selection of suitable materials there are several parameters that must be carefully controlled when fabricating

enhanced as a result of phase separation between the donor and acceptor on meso (> 100 nm) and nanoscales (< 20 nm).(Kim, Cook, Tuladhar, Choulis, Nelson, Durrant, Bradley, Giles,

Relation Between Nanomorphology and Performance of Polymer-Based Solar Cells 129

The effect of thermal annealing on film morhology was clearly demonstrated by bright-field (BF) transmission electron microscopy (TEM) images, recorded in slight defocussing

Fig. 6. Transmission Electron Microscopy (TEM) images show the overview (a) and zoomed in (b), and the corresponding schematic representation (c) of the photoactive layer solar cells.

As can be seen, fibrillar-like P3HT domains (brighter in contrast compared to background) overlap with each other over the whole film. The inset of Fig. 6 shows the selected area electron diffraction (SAED) pattern of the film. The outer ring corresponds to a distance of 0.39 nm, which is typical pi-pi stacking distance of P3HT chains.(Ihn et al., 1993) The crystallinity of P3HT crystals is not very pronounced as seen from the low intensity of the reflection ring. The inner ring in the SAED pattern, corresponding to a d-spacing of 0.46 nm, is seen to be even more diffuse and has been attributed by the nanometer sized PCBM nanocrystals that are homogeneously dispersed throughout the film. Fig. 6 a) shows the BF TEM images of the composite film after annealing (120 C for 60 min). The most pronounced feature in the BF TEM image of the annealed sample is the increased contrast and the appearance of bright fibrillar P3HT crystals throughout the entire film. The width of these crystals remains almost constant compared to the pristine composite film, but on average their length was found to increase over 50 %. The increased crystallinity of P3HT after thermal treatment

Left (a)-(c) images: pristine unannealed P3HT and PCBM blend. Right (a)-(c) images: thermal annealed P3HT and PCBM blend. Insets in (a) figures is the corresponding SAED pattern. The dash line bordered regions represent the extension of existing P3HT crystals in the pristine film or newly developed PCBM-rich domains during the annealing step. The arrow is to indicate the increased intensity of (020) Debye-Scherrer ring from P3HT crystals compared to the SAED pattern shown in the inset of Figure 2a. Reprinted with permission

from (Yang et al., 2005). Copyright 2005 American Chemical Society.

conditions in a P3HT/PCBM blends, as shown in Fig. 6.(Yang et al., 2005)

McCulloch, Ha et al., 2006; Ma et al., 2005)

Fig. 5. Bilayer and bulk-heterojunction solar cells. Reprinted with permission from (Pivrikas et al., 2010a). Copyright 2009, with permission from Elsevier.

BHSC, such as the solution concentration, deposition temperature, donor-acceptor blend ratio, spin speeds using solvents with different boiling points, solvent evaporation kinetics, vapour pressure, solubility, and polarity.(Pivrikas et al., 2010b)

Various methods allowing the control of the film nano-morphology have been introduced in the past.(Chen, Hong, Li & Yang, 2009; Peet et al., 2009) Initially it was observed that the PCE of BHSC significantly increases upon a postproduction treatment, e.g. thermal annealing of solar cells with applied external voltage.(Hoppea & Sariciftci, 2004; Padinger et al., 2003) Other methods used in the past to control the morphology of BHSC involve the use of appropriate solvents with specific boiling point that allow either the increase or decrease of the solvent evaporation rate.(Kim, Choulis, Nelson, Bradley, Cook & Durrant, 2005; Shaheen et al., 2001; Yu et al., 1995) Other methods, such as reducing the drying speed of spin-coated films,(Li et al., 2005; Mihailetchi et al., 2006; Vanlaeke et al., 2006) solubility matching(Troshin et al., 2009) and the melting of bilayers have also been used.(Kim, Liu & Carroll, 2006) It was observed that chemical additives can substitute the post production treatment of BHSC.(Lee et al., 2008) Processing additives are an attractive concept due to the simplicity and suitability for large scale production.

### **3.1 Thermal annealing of devices**

Thermal annealing, by controlling the temperature and annealing time, is typically applied to either the final device or BHJ films in order to improve the nanoscale phase separation between donor and acceptor.(Ma et al., 2005; Sun et al., 2007; Xin et al., 2008; Zhang, Choi, Haliburton, Cleveland, Li, Sun, Ledbetter & Bonner, 2006) Significant improvement in photovoltaic performance after annealing is typically observed in P3HT/PCBM blends.(Padinger et al., 2003) Thermal annealing has the advantage in that it can be applied independently of the film deposition technique. Thermal annealing has also been shown to enhance the crystallinity of the polymer, such as for P3HT, increasing the PCE and the photocurrent due to increased carrier mobility.(Erb et al., 2005) Furthermore, the interconnections between the polymer/fullerene phazes in the interpenetrating network are 8 Will-be-set-by-IN-TECH

Fig. 5. Bilayer and bulk-heterojunction solar cells. Reprinted with permission from (Pivrikas

BHSC, such as the solution concentration, deposition temperature, donor-acceptor blend ratio, spin speeds using solvents with different boiling points, solvent evaporation kinetics, vapour

Various methods allowing the control of the film nano-morphology have been introduced in the past.(Chen, Hong, Li & Yang, 2009; Peet et al., 2009) Initially it was observed that the PCE of BHSC significantly increases upon a postproduction treatment, e.g. thermal annealing of solar cells with applied external voltage.(Hoppea & Sariciftci, 2004; Padinger et al., 2003) Other methods used in the past to control the morphology of BHSC involve the use of appropriate solvents with specific boiling point that allow either the increase or decrease of the solvent evaporation rate.(Kim, Choulis, Nelson, Bradley, Cook & Durrant, 2005; Shaheen et al., 2001; Yu et al., 1995) Other methods, such as reducing the drying speed of spin-coated films,(Li et al., 2005; Mihailetchi et al., 2006; Vanlaeke et al., 2006) solubility matching(Troshin et al., 2009) and the melting of bilayers have also been used.(Kim, Liu & Carroll, 2006) It was observed that chemical additives can substitute the post production treatment of BHSC.(Lee et al., 2008) Processing additives are an attractive concept due to the simplicity and suitability for large

Thermal annealing, by controlling the temperature and annealing time, is typically applied to either the final device or BHJ films in order to improve the nanoscale phase separation between donor and acceptor.(Ma et al., 2005; Sun et al., 2007; Xin et al., 2008; Zhang, Choi, Haliburton, Cleveland, Li, Sun, Ledbetter & Bonner, 2006) Significant improvement in photovoltaic performance after annealing is typically observed in P3HT/PCBM blends.(Padinger et al., 2003) Thermal annealing has the advantage in that it can be applied independently of the film deposition technique. Thermal annealing has also been shown to enhance the crystallinity of the polymer, such as for P3HT, increasing the PCE and the photocurrent due to increased carrier mobility.(Erb et al., 2005) Furthermore, the interconnections between the polymer/fullerene phazes in the interpenetrating network are

et al., 2010a). Copyright 2009, with permission from Elsevier.

pressure, solubility, and polarity.(Pivrikas et al., 2010b)

scale production.

**3.1 Thermal annealing of devices**

enhanced as a result of phase separation between the donor and acceptor on meso (> 100 nm) and nanoscales (< 20 nm).(Kim, Cook, Tuladhar, Choulis, Nelson, Durrant, Bradley, Giles, McCulloch, Ha et al., 2006; Ma et al., 2005)

The effect of thermal annealing on film morhology was clearly demonstrated by bright-field (BF) transmission electron microscopy (TEM) images, recorded in slight defocussing conditions in a P3HT/PCBM blends, as shown in Fig. 6.(Yang et al., 2005)

Fig. 6. Transmission Electron Microscopy (TEM) images show the overview (a) and zoomed in (b), and the corresponding schematic representation (c) of the photoactive layer solar cells. Left (a)-(c) images: pristine unannealed P3HT and PCBM blend. Right (a)-(c) images: thermal annealed P3HT and PCBM blend. Insets in (a) figures is the corresponding SAED pattern. The dash line bordered regions represent the extension of existing P3HT crystals in the pristine film or newly developed PCBM-rich domains during the annealing step. The arrow is to indicate the increased intensity of (020) Debye-Scherrer ring from P3HT crystals compared to the SAED pattern shown in the inset of Figure 2a. Reprinted with permission from (Yang et al., 2005). Copyright 2005 American Chemical Society.

As can be seen, fibrillar-like P3HT domains (brighter in contrast compared to background) overlap with each other over the whole film. The inset of Fig. 6 shows the selected area electron diffraction (SAED) pattern of the film. The outer ring corresponds to a distance of 0.39 nm, which is typical pi-pi stacking distance of P3HT chains.(Ihn et al., 1993) The crystallinity of P3HT crystals is not very pronounced as seen from the low intensity of the reflection ring. The inner ring in the SAED pattern, corresponding to a d-spacing of 0.46 nm, is seen to be even more diffuse and has been attributed by the nanometer sized PCBM nanocrystals that are homogeneously dispersed throughout the film. Fig. 6 a) shows the BF TEM images of the composite film after annealing (120 C for 60 min). The most pronounced feature in the BF TEM image of the annealed sample is the increased contrast and the appearance of bright fibrillar P3HT crystals throughout the entire film. The width of these crystals remains almost constant compared to the pristine composite film, but on average their length was found to increase over 50 %. The increased crystallinity of P3HT after thermal treatment

is oriented parallel - which is the typically observed P3HT orientation. Upon annealing the as-prepared films at various temperatures, the d-spacing along the *a*-axis of the P3HT crystal was found to remain constant, indicating that during the interdiffusion process, the PCBM does not interpenetrate between the side chains of the P3HT crystal structure.(Mayer et al., 2009) The peak width of the diffraction ring, corresponding to the aggregates of PCBM does not change during the interdiffusion process, showing that PCBM remains in an amorphous state with aggregates large enough to scatter incident X-rays. Only a small change in the distribution of P3HT crystal orientations was found to be present at various levels of interdiffusion, while the intensity of the (200) peak of P3HT increased by nearly a factor of two on annealing at 170 C. It was shown that the interdiffusion process has little effect on the crystalline regions of the P3HT film, where the diffusion of PCBM into P3HT occurs within

Relation Between Nanomorphology and Performance of Polymer-Based Solar Cells 131

To determine how interdiffusion within this system affects the growth of the P3HT crystallites, the P3HT crystallite size along the *a*-axis for the bilayer films was compared to pure P3HT films heated under similar conditions (Fig. 7 (f)-(g)). The P3HT crystallite size was estimated using the Scherrer equation and plotted against the fraction of PCBM within the P3HT layer (Fig. 7 (f) ). The crystallite size was found to increase with increasing annealing temperature regardless of the level of interdiffusion. The P3HT crystallite size in the bilayer system was found to increase most rapidly during the first 5 min of annealing, where the crystallite thickness was approching that for a neat P3HT film heated under similar conditions (Fig.

Postproduction treatment requires a rather well controlled environment, it adds an additional fabrication costs to the solar cell manufacturing process, which might not be attractive for large-scale industrial production. Furthermore, some material systems, like the low band gap organic semiconductor poly[2,6-(4,4-bis-(2-ethylhexyl)-4H-cyclopenta[2,1-b;3,4-b0] dithiophene)-alt-4,7-(2,1,3-benzothiadiazole)] (PCPDTBT) blended with [6,6]-phenyl C71-butyric acid methyl ester (C71-PCBM), do not shown any improvement upon thermal

Phase separation and molecular self-organization can be influenced by solvent evaporation since the solvent establishes the film evolution environment. Slow drying or solvent annealing techniques have also been used to control the morphology of the blends by changing the rate of solvent removal.(Li et al., 2005; Li, Yao, Yang, Shrotriya, Yang & Yang, 2007; Sivula et al., 2006) The use of different solvents and their effect on the film nano-structure of BHSC has been studied in detail in the past.(Li, Shrotriya, Yao, Huang & Yang, 2007) High boiling point solvents were used with the device placed in an enclosed container, in which the atmosphere

Grazing-incidence x-ray diffraction (GIXRD) studies provided evidence that the solvent evaporation rate directly influences the polymer chain arrangement in the film.(Chu et al., 2008) It was shown that the use of higher boiling point solvent strongly improves the PCE of MDMO-PPV and PCBM blends.(Shaheen et al., 2001) Higher PCE values due to improved film morphology and crystallinity have been reached by substituting chloroform with chlorobenzene for P3HT/PCBM BHSC.(Ma et al., 2005) The difference between chlorobenzene and 1,2-dichloro benzene for use as a solvent was shown in the novel low bandgap polymer PFco-DTB and C71-PCBM blend systems, where chlorobenzene resulted in films with higher

the disordered regions of P3HT.

rapidly saturates with the solvent.

7 (g) ).

annealing.

**3.2 Solvent effects**

is visible from the increased intensity of the reflection ring in the SAED pattern (inset of Fig. 6 a)). Larger and darker PCBM rich areas can be observed suggesting an increased phase demixing between P3HT and PCBM. It was concluded that the crystallinity of P3HT is improved upon annealing and the demixing between the two components is increased, but large-scale phase separation does not occur. The resulting interpenetrating networks composed of P3HT crystals with a high aspect ratio and aggregated nanocrystalline PCBM domains provide continuous pathways in the entire photoactive layer for efficient hole and electron transport.

In order to further understand the extent of thermal annealing, 2-D X-ray scattering in a grazing incidence geometry (GIWAXS) was used to study the development of the crystalline structure of P3HT and PCBM during the interdiffusion process at various temperatures. 2D GIWAXS patterns of as-prepared P3HT/PCBM bilayers and annealed samples are shown in Fig. 7.(Treat et al., 2011) The diffraction patterns for P3HT shows that the *a*-axis of the P3HT crystals is predominantly oriented perpendicular to the substrate and the *b*-axis ( pi-stacking)

Fig. 7. Two-dimensional GIWAXS of a P3HT/PCBM bilayer on Si a) as-cast and annealed at b) 70 C, c) 110 C, d) 150 C, and e) 170 C for 5 min. f) The Scherrer equation was used to extract the P3HT crystallite thickness along the a- axis from the full-width-at-half-maximum of the (100) reflection. PCBM incorporation from the DSIMS measurements was plotted for comparison at various annealing temperatures. g) Growth in the crystal thickness with time using in-situ heating 2D GIWAXS of a P3HT/PCBM bilayer on Si at 110 C (orange) and 170 C (blue). The Scherrer equation was used to determine crystal thickness from the (100) reflection corresponding to P3HT. The dotted line corresponds to a neat P3HT/Si sample heated for 5 min at 110 C (orange) and 170 C (blue). Copyright 2011 Wiley. Used with permission from (Treat et al., 2011).

is oriented parallel - which is the typically observed P3HT orientation. Upon annealing the as-prepared films at various temperatures, the d-spacing along the *a*-axis of the P3HT crystal was found to remain constant, indicating that during the interdiffusion process, the PCBM does not interpenetrate between the side chains of the P3HT crystal structure.(Mayer et al., 2009) The peak width of the diffraction ring, corresponding to the aggregates of PCBM does not change during the interdiffusion process, showing that PCBM remains in an amorphous state with aggregates large enough to scatter incident X-rays. Only a small change in the distribution of P3HT crystal orientations was found to be present at various levels of interdiffusion, while the intensity of the (200) peak of P3HT increased by nearly a factor of two on annealing at 170 C. It was shown that the interdiffusion process has little effect on the crystalline regions of the P3HT film, where the diffusion of PCBM into P3HT occurs within the disordered regions of P3HT.

To determine how interdiffusion within this system affects the growth of the P3HT crystallites, the P3HT crystallite size along the *a*-axis for the bilayer films was compared to pure P3HT films heated under similar conditions (Fig. 7 (f)-(g)). The P3HT crystallite size was estimated using the Scherrer equation and plotted against the fraction of PCBM within the P3HT layer (Fig. 7 (f) ). The crystallite size was found to increase with increasing annealing temperature regardless of the level of interdiffusion. The P3HT crystallite size in the bilayer system was found to increase most rapidly during the first 5 min of annealing, where the crystallite thickness was approching that for a neat P3HT film heated under similar conditions (Fig. 7 (g) ).
