Efficient Phthalimide CopolymerBased Bulk ... - Wiley Online Library

7 downloads 1891 Views 1MB Size Report
Mar 13, 2012 - E-mail: [email protected]. DOI: 10.1002/aenm.201100718. 1. Introduction. The two-phase morphology of bulk heterojunction (BHJ) pol-.
www.advenergymat.de www.MaterialsViews.com

FULL PAPER

Efficient Phthalimide Copolymer-Based Bulk Heterojunction Solar Cells: How the Processing Additive Influences Nanoscale Morphology and Photovoltaic Properties Hao Xin, Xugang Guo, Guoqiang Ren, Mark D. Watson,* and Samson A. Jenekhe*

mers solar cells.[11,13,22] Initial investigation suggested that the role of the processing The power conversion efficiency of poly(N-(2-ethylhexyl)-3,6-bis(4-dodecyadditive in controlling the two-phase morloxythiophen-2-yl)phthalimide) (PhBTEH)/fullerene bulk heterojunction solar phology was to increase the domain size of cells improves from 0.43 to 4.1% by using a processing additive. The underthe polymer[15] and/or the fullerene.[13] The lying mechanism for the almost 10-fold enhancement in solar cell performdomain size of the polymer was controlled ance is found to be inhibition of fullerene intercalation into the polymer side by enhancing the polymer aggregation chains and regulation of the relative crystallization/aggregation rates of the during the spin-coating and drying while delaying the aggregation of the fullerene polymer and fullerene. An optimal interconnected two-phase morphology by virtue of its selective solubility in the with 15–20 nm domains is obtained when a processing additive is used higher-boiling additive. However, more compared with 100–300 nm domains without the additive. The results demrecent studies indicate that the processing onstrate that a processing additive provides an effective means of controlling additive improves performance of polymer both the fullerene intercalation in polymer/fullerene blends and the domain solar cells by reducing the domain size of the phase separation and creating a more sizes of their phase-separated nanoscale morphology. uniform BHJ morphology.[23,24] Details of how the processing additive influences the phase separation, the domain sizes of the two-phase mor1. Introduction phology, and the photovoltaic properties of the BHJ as well as applicability of the method to other D–A conjugated copolymer The two-phase morphology of bulk heterojunction (BHJ) polfilms remain unknown. ymer solar cells is critical to the performance of the devices.[1–16] The complex nature of the morphology of polymer/fullerene Many methods have thus been explored towards achieving a BHJ blend films and its evolution during and after deposition nanoscale bicontinuous morphology, including thermal[17,18] of the BHJ films has been highlighted by the recent discovery and solvent annealing[6,19] and self-assembly of polymer of a fullerene-intercalated polymer phase in which the fullerene nanowires.[7,8,10] These strategies have been effective to varying molecules are intercalated into the side chains of the conjugated degrees for controlling the morphology and performance of polymer and can be thought of as a ‘co-crystal’ phase.[25–27] Such BHJ devices based on poly(3-alkylthiophene)s but are either a fullerene-intercalated polymer phase while useful for creating less effective or do not work at all for many other polymer semicharges and increasing the mobility of holes in the polymer as conductors.[11,20,21] In the case of some donor–acceptor (D–A) observed in the poly(p-phenylene vinylene derivative MDMOcopolymers, it has been found that processing the polymer/ PPV,[27] is undesirable for high performance solar cells for fullerene blend film from a binary solvent mixture containing a various reasons. It could preclude or reduce the formation of small amount of a higher-boiling-point additive, which is a good well-defined, separate hole-conducting polymer and electronsolvent for the fullerene but a bad solvent for the polymer, can conducting fullerene domains within the BHJ film, lowering significantly improve the photovoltaic properties of the polycharge photogeneration, charge transport, and ultimately reducing the photovoltaic efficiency. Existence of a fullereneintercalated polymer phase appears to account for the large Dr. H. Xin, G. Ren, Prof. S. A. Jenekhe excess of fullerene loading (e.g., polymer:fullerene = 1:3 to 1:4) Department of Chemical Engineering and needed to achieve good device performance in some polymer/ Department of Chemistry, University of Washington Box 351750, Seattle, Washington 98195-1750, USA fullerene systems.[25–27] In addition, the close intermolecular E-mail: [email protected] contacts in the fullerene-intercalated polymer phase could facilDr. X. Guo, Prof. M. D. Watson itate the formation of exciplexes and undesirable enhancement Department of Chemistry of charge recombination.[28–30] By using a fullerene derivative University of Kentucky that is much larger than [6,6]-phenyl-C71-butyric acid methyl Lexington, Kentucky 40506-0055, USA ester (PC71BM), intercalation into the side chains of the polymer, E-mail: [email protected] poly(2,5-bis(3-hexadecylthiophen-2-yl)thieno[3,2-b]thiophene DOI: 10.1002/aenm.201100718 (pBTTT), could be completely prevented.[26] However, how

Adv. Energy Mater. 2012, 2, 575–582

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

575

www.advenergymat.de

FULL PAPER

www.MaterialsViews.com

b

0.8

Without ODT, SD Without ODT, FD With ODT, SD With ODT, FD

5

Absorbance / arb. units

Current density / mA cm-2

10

Without ODT, SD Without ODT, FD With ODT, SD With ODT, FD

0.6

0

0.4

-5

0.2

-10 -15

c

-0.4

-0.2

0

0.2

Voltage / V

0.4

0.6

0.8

0 300

400

500

600

700

Wavelength / nm

800

Figure 1. a) Molecular structures of PhBTEH and PC71BM. b) The J/V curves of PhBTEH:PC71BM (1:1) blend solar cells from different film processing conditions: slow drying (SD); fast drying (FD). c) Absorption spectra measured directly from the devices in (b); ITO/PEDOT:PSS substrate was used as a reference without normalization.

to prevent intercalation of conventional fullerenes (PC71BM, PC61BM) into the side chains of polymer semiconductors in BHJ solar cells remains to be addressed. In this paper, we report the results of an investigation of the above related questions, i.e. how does a processing additive influence the photovoltaic properties of D–A copolymer semiconductor-based BHJ solar cells and how can the intercalation of conventional fullerenes into the side chains of a D–A copolymer in BHJ solar cells be prevented. We used a combination of photovoltaic measurements, transmission electron microscopy (TEM) and atomic force microscopy (AFM) imaging, and X-ray diffraction of BHJ blends of poly(N-(2ethylhexyl)-3,6-bis(4-dodecyloxythiophen-2-yl)phthalimide) (PhBTEH) and PC71BM to show that devices processed without an additive have a poor conversion efficiency (0.43%) and a morphology characterized by large domains (100–300 nm) and a fullerene-intercalated polymer phase. We also show that by employing a processing additive in the fabrication of the BHJ solar cells from PhBTEH:PC71BM blends the power conversion efficiency is enhanced to 4.1% while the morphology is characterized by 15–20 nm domains completely free of a fullereneintercalated polymer phase. Our results thus demonstrate that the processing additive strategy can be beneficially applied to the new D–A copolymer PhBTEH and that its underlying 576

wileyonlinelibrary.com

mechanism involves the inhibition of fullerene intercalation into the polymer side chains and the reduction of the polymer and fullerene domain sizes.

2. Results and Discussion The molecular structures of the donor polymer, (PhBTEH), and the fullerene derivative, [6,6]-phenyl-C71-butyric acid methyl ester (PC71BM), used in this study are shown in Figure 1a. We have previously reported the synthesis and field-effect charge transport properties of PhBTEH and its derivative, poly(N(dodecyl)-3,6-bis(4-dodecyloxythiophen-2-yl)phthalimide) (PhBT12).[31] However, the photovoltaic properties of PhBTEH have not been reported. The processing solvent used in this study was 1,2-dichlorobenzene (ODCB) and the additive was 1,8-octanedithiol (ODT). The ratio of ODCB:ODT was optimized at 100:3 (vol:vol) in the binary solvent mixture, which gave the best solar cell performance under optimized film drying conditions. The composition of PhBTEH:PC71BM blends was fixed at 1:1 (wt:wt). The active layers were spincoated from warm (80 °C) solutions and had a thickness of 80 nm. The spin-coating was performed in an argon-filled glovebox.

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Adv. Energy Mater. 2012, 2, 575–582

www.advenergymat.de www.MaterialsViews.com

FULL PAPER

Table 1. Photovoltaic Properties of PhBTEH:PC71BM (1:1 wt:wt) Solar Cells. Solvent

Drying Condition

μh-SCLC [cm2 V−1s−1]

ODCB

Slow dry

ODCB

Jsc [mA cm−2]

Voc [V]

FF

PCEa) [%]

4.27 × 10−6

1.63

0.55

0.48

0.43 (0.45)

Fast dry

2.93 × 10−4

3.66

0.58

0.55

1.16 (1.22)

ODCB+ODT

Slow dry

2.35 × 10−4

8.99

0.53

0.60

2.85 (2.87)

ODCB+ODT

Fast dry

3.55 × 10−4

12.5

0.54

0.61

4.09 (4.12)

a)The

numbers in parentheses are the highest efficiency.

It can be expected that the domain sizes and the nature of the phases of the BHJ morphology depend on the kinetics of the phase separation process and the relative crystallization/ self-assembly rates of the polymer and fullerene. Because of the selective solubility of the fullerene in the additive based on our results, we hypothesized that the additive can likely also inhibit the fullerene intercalation into the polymer side chains during the phase separation process. To control the overall phase segregation kinetics, we used two drying conditions or solvent evaporation rates from the spin coated BHJ blend film: a) slow; b) fast. Slow drying (SD) was intended to allow the polymer and fullerene phases enough time to grow and involved placing the spin-coated BHJ film in a covered petri dish at room temperature for 60 min. In contrast, fast drying (FD) aimed to freeze the phase-segregated morphology in the early stages and involved placing the spin-coated BHJ film in the small vacuum chamber of the glovebox at room temperature (25 °C) for 30 min. Figure 1b shows representative current density–voltage (J–V) curves for the four PhBTEH solar cells processed without and with ODT additive and dried under the slow and fast conditions. The solar cell parameters are summarized in Table 1. Without using a processing additive, the average power conversion efficiency (PCE) was rather poor at 0.43 ± 0.02% for the SD blend films; the performance improved to 1.2 ± 0.05% in the FD blend films. In the case of devices prepared using the ODT processing additive and SD, the observed PCE was 2.9 ± 0.02% with a Jsc of 8.99 mA cm−2, a Voc of 0.55 V, and FF of 0.48. Devices made by using a processing additive in conjunction with rapid film drying under vacuum, have a significantly increased efficiency of 4.1 ± 0.03% PCE. The maximum efficiency achieved in the best solar cell was 4.12% with Jsc, Voc, and FF of 12.4 mA cm−2, 0.54 V, and 0.62, respectively. The overall enhancement in performance of the PhBTEH solar cells has two sources to be discussed in detail below: the processing additive per se and the fast drying. Overall, the photovoltaic efficiency is enhanced by a factor of 9.5 in going from a slowly dried BHJ film without using a processing additive to rapidly dried film processed with ODT additive. All the devices showed a very similar Voc of 0.53/boxh 0.55 V, indicating that the film processing conditions do not affect the Voc of PhBTEH/fullerene solar cells. The most significant parameter affecting the solar cell performance is the Jsc. The Jsc is improved by the processing additive, respectively, by a factor of 5.5 (from 1.63 to 8.99 mA cm−2) and 3.4 (from 3.66 to 12.5 mA cm−2) in the SD and FD films. In addition, we note that with or without using a processing additive, the FD films

Adv. Energy Mater. 2012, 2, 575–582

Figure 2. TEM images of PhBTEH:PC71BM (1:1) blend films processed under different conditions. a) Without additive and slow drying; b) without additive and fast drying; c) with additive and slow drying; d) with additive and fast drying. The scale bars are 0.5 μm in (a) and (b) and 200 nm in (c) and (d).

gave a higher Jsc than the SD films, increasing from 1.63 to 3.66 mA cm−2 and 8.99 to 12.5 mA cm−2 respectively without and with a processing additive. The fill factor FF is also greatly changed by the BHJ film processing conditions. By using a processing additive the FF could be increased from 0.48 to 0.60 in SD films and from 0.55 to 0.61 in FD films. Considering that all the BHJ films have the same polymer:fullerene composition and film thickness, the large difference in the photovoltaic performance must come from the morphology, which is analyzed and discussed below. Figure 1c shows the absorption spectra of PhBTEH:PC71BM blend films measured directly on the solar cell devices. Significant absorption enhancement in the 600–715 nm wavelength range with distinct peaks at 625 and 690 nm, characteristic of the π-stacked polymer film,[31] were observed in the films processed via ODT additive, indicating higher order of the polymer domains in the PhBTEH:PC71BM films. The increased absorption in the 600–715 nm region likely also contributes to the observed improved Jsc in the BHJ devices processed by means of ODT additive. TEM images collected directly from the PhBTEH:PC71BM (1:1, wt:wt) solar cell films processed under the four conditions are shown in Figure 2. Large scale phase separation with a continuous darker phase embedded with discrete micrometer-size light domains is seen in SD films processed without ODT additive (Figure 2a). X-ray diffraction (Figure 3a) of the sample corresponding to the image of Figure 2a revealed a broad diffraction maximum at 2θ = 5.38°, or a d-spacing of 1.64 nm, which cannot be assigned to any diffraction from the pure polymer

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

577

www.advenergymat.de

FULL PAPER

www.MaterialsViews.com

Intensity / arb. units

a

b

Spin-coated blend films No ODT, slow dry No ODT, quick dry With ODT, slow dry With ODT, quick dry

(200)' (100)

c

Drop casted blend films (100)

No ODT With ODT (200)' (200)

3

4

5 6 7 8 2-Theta / degree

9

10

Figure 3. a) XRD patterns of spin-coated PhBTEH:PC71BM (1:1) blend films (top, measured on solar cell devices) and drop-casted films (bottom) under different processing conditions indicated in the legend. b) Illustration of PhBTEH domain in blend film. c) Illustration of PhBTEH:PC71BM blend film with intercalated fullerene. The (100) d-spacing of the polymer lamellar is expanded by 0.9 nm after intercalation of PC71BM into the polymer side chains.

ODT additive (Table 1). The low mobility of holes and the lack of electron percolation due to the formation of the fullereneintercalated polymer phase can account for the very poor photovoltaic efficiency of the slowly dried devices processed without an additive.[26] To further confirm the intercalation of the fullerene in PhBTEH, X-ray diffraction of PhBTEH:PC71BM blend films with different mole ratios was performed and the results are shown in Figure 4a. The XRD patterns of pure polymer films are also shown for comparison. Because a perfect intercalation requires 1:1 mole ratio between the polymer repeating unit and fullerene molecule, by decreasing or increasing the fullerene loading, one should see a change in the diffraction between the pure polymer phase and the intercalation phase. Indeed, in the 1:0.5 blend film with less fullerene loading, except for diffraction peak at 2θ = 5.38° from the intercalation phase, a strong diffraction peak at 2θ = 3.36° from the pure polymer phase also appeared. By increasing the fullerene loading from 1:0.5 to 1:1 and 1:4, the relative diffraction intensity of the intercalation phase gradually increased while the intensity of the pure polymer phase decreased, confirming that most of the polymer chains are involved in the intercalation. The appearance of diffraction from the pure polymer in the diffraction patterns of the 1:1 blend and 1:4 blend films indicates that the intercalation is not perfect and that defects exist in both films. However, considering that the peak at 2θ = 5.38° is the second order of the diffraction from the intercalation phase, the fraction of the pure polymer phase in the 1:4 blend film is very small. Nevertheless, these results confirm the intercalation of the fullerene into the side-chains of PhBTEH in PhBTEH:PC71BM films. We note that the polymer shows stronger diffraction in the blend films with fullerene than as a pure film, although the reason for this remains unclear. After thermal annealing, the intercalation phase disappeared in all the films along with the intensity enhancement of the pure polymer peak in the XRD pattern (Figure 4a), indicating that the extraction of the fullerene from the polymer side chains had occurred upon heat treatment. TEM imaging of FD BHJ film processed without ODT additive (Figure 2b) shows an interconnected polymer network

578

wileyonlinelibrary.com

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Intensity / arb. units

Intensity / arb. units

phase.[31] The diffraction peak at 2θ = 5.38° was also observed in a drop-cast PhBTEH:PC71BM (1:1) blend film along with the lamellar (100) peak of the pure PhBTEH phase at 2θ = 3.36° or a d-spacing of 2.4 nm.[31] The new diffraction peak at 2θ = 5.38° may be assigned to the second order diffraction (200)’ of the fullerene-intercalated polymer phase.[24] Illustration of the lamellar PhBTEH structure and the corresponding intercalated PhBTEH:PC71BM lamellar structure are shown in Figure 3a and 3b, respectively. The d-spacing calculated from the (200)’ diffraction peak is 3.28 nm, which is expanded by 0.9 nm after fullerene intercalation, in good agreement with McGehee’s results in other polymer/fullerene intercalation systems.[25] According to McGehee’s calculation,[25] in a perfect intercalation phase, the molar ratio of the polymer repeating unit to the fullerene should be 1 to a (100) 1; the 1:1 weight ratio of PhBTEH:PC71BM b in this study corresponds to a molar ratio of (100) After annealing 1.26:1, suggesting that there is extra polymer After annealing 1:4 1:4 that may form a pure PhBTEH phase. The 1:1 1:1 observed (100) diffraction in the drop-casted 1:0.5 1:0 PhBTEH:PC71BM blend film supports the 1:0.5 existence of a pure polymer phase, in agreeAs drop cast (200)' (100) As spin-coated (200)' ment with what is observed in the TEM 1:4 1:4 imaging (Figure 2a). The absence of a (100) 1:1 1:1 1:0.5 diffraction in the spin-coated blend film can 1:0.5 1:0 be explained by the very thin film (80 nm) and the discontinuous small domains of the 2 3 4 5 6 7 8 9 10 2 3 4 5 6 7 8 9 10 polymer phase. Indeed, the hole mobility 2-Theta / degree 2-Theta / degree of this film measured by the space charge Figure 4. X-ray diffractions patterns of a) drop cast and b) spin-coated and fast dried limited current (SCLC) is only 4.27 × PhBTEH:PC71BM blend films with mole ratio of 1:0.5, 1:1, and 1:4 before and after thermal 10−6 cm2 V−1 s−1, almost two orders of mag- annealing. Pure PhBTEH films are also shown in (a) for comparison. Films in (a) were drop nitude lower than the pure PhBTEH film and cast on quartz substrates and films in (b) were spin-coacted on ITO/PEDOT substrates. The PhBTEH:PC71BM blend films processed with annealing temperature and time were 170 °C and 30 min for all films.

Adv. Energy Mater. 2012, 2, 575–582

www.advenergymat.de

Adv. Energy Mater. 2012, 2, 575–582

1.2 1

FULL PAPER

with uniformly dispersed dark phase with domain sizes of 100–300 nm. The interconnected polymer network explains the increased SCLC hole mobility (2.93 × 10−4 cm2 V−1 s−1, Table 1) and the improved photovoltaic efficiency (1.2 ± 0.05% PCE) relative to the slowly dried BHJ films. Although a diffraction peak from the intercalated phase was not detected in the XRD pattern (Figure 3a) due to very thin film and less ordered structure, we believe that the dark areas in the image of Figure 2b are mainly due to the intercalated phase. Partial evidence for this is that both films show very similar absorption spectra (Figure 1c). The high order of the intercalation phase and appearance of the isolated pure polymer domains in SD films (Figure 3a) can be understood from the longer residence time of the solvent in the film which allowed polymer and fullerene enough time to move and organize. To prove the existence of intercalation in FD blend films processed without ODT, XRD of PhBTEH:PC71BM blend films with a molar ratio of 1:0.5, 1:1, and 1:4 before and after thermal annealing were measured and the results are shown in Figure 4b. To enhance the diffraction signal, these films are slightly thicker (∼120 nm) than those in solar cell devices (80 nm). Although very weak, the diffraction peak from the intercalation phase was detected in all three films, confirming the existence of intercalation in films processed without ODT. In addition, for the 1:0.5 blend film, similar to the slowly dried film (Figure 4a), the diffraction peak from the pure polymer was also observed, indicating that highly crystalline polymer domains can be formed even under fast drying (FD) condition. However, the pure polymer phase did not appear in both 1:1 and 1:4 blend films, indicating that such a phase was not formed. After thermal annealing, distinct polymer diffraction appeared in all films along with the disappearance of the diffraction from the intercalation phase. These results further confirm the intercalation in the fast dried films. We note that when PC71BM was replaced by a larger fullerene derivative indene-C60 bisadduct (ICBA) (Supporting Information, SI, Scheme S1), only diffraction peak from the pure polymer phase was observed in both slow and fast dried PhBTEH/ICBA (1:1) blend films (SI, Figure S1), indicating the prevention of intercalation, similar to McGhee’s results on a large fullerene (bisPC71BM).[26] A morphology of severe phase separation with large fullerene aggregates was seen in both slow and fast dried PhBTEH/ICBA (1:1) blend films (SI, Figure S2). The morphology of both SD and FD BHJ films using ODT additive revealed polymer domains with widths of only 15–20 nm (Figure 2c,d). That the polymer domains have a higher degree of order was confirmed by the appearance of the (100) lamellar diffraction in the corresponding XRD patterns (Figure 3a) and the enhanced absorption in the 600–700 nm region (Figure 1c). The absence of the (200)’ peak in both spin-coated and drop-casted films reveals that the fullerene-intercalated phase was absent in these films. We note that the polymer domains in quickly and slowly dried films are very similar in size, indicating that the polymer self-assembly is very fast when the processing additive is used. The difference between the morphology of the two films is the length of the polymer fibers and the size of the fullerene domains. In the SD film, the polymer fibers have a length of about 500 nm while the fullerene aggregates are of micrometer sizes (Figure 2c),

Absorbance / arb. units

www.MaterialsViews.com

Solution in ODCB Without ODT, as spin-coated film Without ODT, dry film With ODT, as spin-coated film With ODT, dry film

0.8 0.6 0.4 0.2 0

400 450 500 550 600 650 700 750 800

Wavelength / nm Figure 5. Absorption spectra of PhBTEH in ODCB solution and the as spin-coated and dried PhBTEH films processed with and without ODT additive.

resulting in a very nonuniform film. Large fullerene aggregates and polymer domains with bundled polymer fibers were also observed in the surface morphology of the SD film via AFM imaging (SI, Figure S3). The surface roughness of this film is 14.8 nm. In contrast, the SD film consisted of 100 nm long and 15–20 nm wide polymer nanofibers and similar-sized fullerene domains (Figure 2d), an ideal morphology for exciton dissociation and charge carrier transport. AFM imaging (SI, Figure S3) revealed a very smooth surface with a roughness of only 1.45 nm and no fibers present on the surface. To further understand the role of the processing additive on the self-assembly of the polymer, pure PhBTEH films were prepared with and without ODT additive. The absorption spectra and morphology (TEM and AFM images) of these films are shown in Figure 5 and Figure 6, respectively. For consistency, these films were spin-coated on top of PEDOT/ITO substrates under conditions identical to those used in the solar cells (e.g., 80 °C solutions, films dried in vacuum). With or without ODT additive, the freshly spin-coated, non-dried films have absorption spectra that are identical to the solution absorption spectrum, indicating that the final mode of self-assembly is slow on this time scale and not achieved until the drying process. After drying, the film processed with ODT additive shows about 1.5 times higher absorption intensity and a more prominent characteristic peak at 700 nm, compared to the film obtained without an additive, indicating that a higher crystallinity is obtained by using an additive. The higher degree of molecular order in the additive-processed film is also confirmed by the appearance of the (100) lamellar diffraction peak in the XRD diffraction pattern which was absent from the film processed without an additive (SI, Figure S4). The morphology of the two films revealed by imaging is also very different (Figure 5). A smooth and featureless surface with a roughness of 2.05 nm is seen in the film

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

579

www.advenergymat.de

FULL PAPER

www.MaterialsViews.com

Figure 6. TEM (top) and AFM (bottom) images of pure PhBTEH films obtained without (a,b,e) and with (c,d,f) a processing additive. Images (b) and (d) are the magnified images of (a) and (c), respectively. The dimension of the AFM images is 5 × 5 μm and the height is 50 nm.

obtained without a processing additive (Figure 5a,b,e). In contrast, ordered cylinder-like domains with sizes of 200–300 nm and a height of about 40–50 nm are seen in the surface of films processed with additive (Figure 5f). These domains are composed of dense nanofibers with a width of 15–20 nm (Figure 5d), similar to those observed in the blend film (Figure 2d). The surface roughness of this film is 14.1 nm. We conclude that the additive can induce fast crystallization of PhBTEH and thereby regulate the domain size of the polymer. Fullerene intercalation can occur in any polymer/fullerene systems if the polymer side chains have enough space to accommodate the fullerene molecules even if the resulting intercalated fullerene/polymer complex is not sufficiently crystalline to diffract X-ray.[25] Our above results on PhBTEH:fullerene systems demonstrated that the use of a processing additive prevents intercalation and thereby enhances the photovoltaic efficiency. To test the generality of the additive approach to preventing intercalation, we have also briefly investigated poly (3,3´˝dialkylquaterthiophene) (PQT-12), which is known to form intercalation phases with fullerene.[25] We fabricated PQT12:PC71BM solar cells under identical FD conditions to that of PhBETH:PC71BM solar cells (SI, Table S1 and Figure S5). We found that without using ODT processing additive, 1:1 (wt:wt) PQT-12:PC71BM solar cells, which presumably contain the intercalated phase,[25] gave a low power conversion efficiency of only 0.38%. By increasing the fullerene loading to achieve electron percolation, the efficiency in PQT-12:PC71BM (1:4 wt:wt) solar cells was improved to 0.54% (SI, Table S1 and Figure S5). After using ODT processing additive, the efficiency of the 1:1 blend film was improved to 1.2% PCE, a factor of 3.7 enhancement. In contrast, the observed efficiency in PQT-12:PC71BM (1:4 wt:wt) solar cells processed with ODT additive was only 580

wileyonlinelibrary.com

1.0%. These results further demonstrate that a processing additive can prevent intercalation in polymer:fullerene systems and achieve better photovoltaic performance with a normal (1:1 wt:wt) fullerene loading. The state-of-the-art p-type semiconductors for highly efficiency BHJ solar cells are generally donor-acceptor copolymers that have either different side chains on the donor and acceptor counits or side chains on only the donor or acceptor moieties.[14,24,32–34] To the extent that the use of processing additives enhances the photovoltaic efficiency of these polymers in BHJ blends with fullerenes, intercalation phenomena are to be expected in such systems. The underlying mechanism proposed here, i.e., inhibition of fullerene intercalation into polymer side chains and regulation of the relative aggregation/crystallization of the polymer and fullerene components, provides new insights on the morphology and performance of BHJ solar cells.

3. Conclusion We have used the donor-acceptor copolymer PhBTEH to investigate how a processing additive and film-drying rate enhance the photovoltaic performance of polymer/fullerene bulk heterojunction solar cells. The observed enhancement of the power conversion efficiency from 0.43 to 4.1% by using a processing additive was found to be a result of inhibition of fullerene (PC71BM) intercalation into the PhBTEH side chains and reduction of the phase segregated polymer/fullerene domains to 15–20 nm. A processing additive is thus a highly effective way of preventing the formation of undesired fullerene-intercalated polymer phases while regulating the nanoscale morphology and composition of bulk heterojunction polymer solar cells.

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Adv. Energy Mater. 2012, 2, 575–582

www.advenergymat.de www.MaterialsViews.com

Materials: The [6,6]-phenyl-C71-butyric acid methyl ester (PC71BM, > 99.5%) was obtained from American Dye Source, Inc. (Quebec, Canada). Poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) (Baytron P VP AI 4083) was purchased from H. C. Stark (Newton, MA) and passed through a 0.45 μm filter before spin-coating. All chemicals were used as received without further purification. The synthesis of PhBTEH (Mn = 117.3 kDa) was previously reported.[31] A PhBTEH solution (10 mg mL−1) was prepared by dissolving 20 mg sample in 2 mL 1,2-dichlorobenzene (ODCB) on a 100 °C hot plate with stirring and the hot solution was filtered through a 0.45 μm GHP filter. PC71BM solution (60 mg mL−1) was dissolved in the same solvent by stirring overnight at 40 °C and passed through a 0.2 μm GHP filter. A PhBTEH:fullerene (1:1) blend solution was prepared by mixing PhBTEH and PC71BM solutions at a ratio of 1:6 (v:v). For solutions with a processing additive (ODT), a small amount of ODT (ODT:ODCB = 3:100) was added to PhBTEH:fullerene blend solution. Both solutions were heated at 80 °C for 20 min before spin-coating. The solutions were prepared and stored in a glove box. Device Fabrication: Solar cells were fabricated by first spin-coating a PEDOT buffer layer on top of ITO-coated glass substrates (10 Ω −1, Shanghai B. Tree Tech. Consult Co., Ltd, Shanghai, China) at 3500 rpm for 40 s and annealing at 150 °C for 10 min under vacuum. The thickness of PEDOT was around 30 nm. The active layers were spin-coated on top of the PEDOT in a glove box at a speed of 860 rpm for 30 s. After deposition of the active layer, the substrates were either kept in the petri dishes for 60 min (slow drying) or immediately put in the connecting chamber of the glove box for fast drying. The substrates were then taken out of the glove box and loaded in a thermal evaporator for cathode deposition. The cathode, consisting of 1.0 nm LiF and 100 nm Al layers, was sequentially deposited through a shadow mask on top of the active layers after the chamber vacuum reached 8 × 10−7 torr. Each substrate contained 5 pixels, each with an active area of 4.0 mm2. Devices for space-charge-limited current (SCLC) measurement were fabricated in a manner similar to the solar cells. The main difference is that in order to facilitate hole-only injection and transport, gold electrodes were deposited instead of the LiF/Al cathode used in the solar cells. Characterizations: Film thickness was measured on an Alpha-Step 500 profilometer (KLA-Tencor, San Jose, CA). Current–voltage characteristics of both solar cells and SCLC devices were measured by using a HP4155A semiconductor parameter analyzer (Yokogawa Hewlett-Packard, Tokyo). The light intensity of AM 1.5 sunlight from a filtered Xe lamp was calibrated by a Si photodiode calibrated at the National Renewable Energy Laboratory (NREL). The SCLC characteristics were measured under dark conditions. All the characterization steps were carried out under ambient laboratory air. UV–vis absorption spectra were recorded on a Perkin-Elmer model Lambda 900 UV–vis–NIR spectrophotometer. Bright-field transmission electron images (BF-TEM) were measured on An FEI Tecnai G2 F20 TEM at 200 kV accelerating voltage. The images were slightly defocused to enhance the phase contrast and were then acquired with a CCD camera and recorded with Gatan DigitalMicrograph software with proper exposure time. Atomic force microscopy (AFM) images were measured from the same films as in the solar cell devices using Dimension 3100 Scanning Probe Microscope (Veeco Instruments Inc., Woodbury, NY. Diffraction data were collected from a Bruker-AXS D8 Focus diffractometer with Cu Kα beam (40 kV, 40 mA).

Supporting Information Supporting Information is available from the Wiley Online Library or from the author.

Acknowledgements This report is based on work (Excitonic Solar Cells) supported by the US Department of Energy, Office of Basic Energy Sciences, Division

Adv. Energy Mater. 2012, 2, 575–582

FULL PAPER

4. Experimental Section

of Materials Sciences under Award No. DE-FG02-07ER46467. SAJ also acknowledges the support from NSF (DMR-0805259) for the study of donor-acceptor conjugated copolymers. Polymer synthesis at the University of Kentucky was supported by the NSF (CHE-0616759). Part of this work was conducted at the University of Washington NanoTech User Facility, a member of the NSF National Nanotechnology Infrastructure Network (NNIN). Received: November 26, 2011 Published online: March 13, 2012

[1] S. Gunes, H. Neugebauer, N. S. Sariciftci, Chem. Rev. 2007, 107, 1324. [2] P. L. T. Boudreault, A. Najari, M. Leclerc, Chem. Mater. 2011, 23, 456. [3] B. C. Thompson, J. M. J. Frechet, Angew. Chem., Int. Ed. 2008, 47, 58. [4] X. N. Yang, J. Loos, S. C. Veenstra, W. J. H. Verhees, M. M. Wienk, J. M. Kroon, M. A. J. Michels, R. A. J. Janssen, Nano Lett. 2005, 5, 579. [5] A. R. Campbell, J. M. Hodgkiss, S. Westenhoff, I. A. Howard, R. A. Marsh, C. R. McNeill, R. H. Friend, N. C. Greenham, Nano Lett. 2008, 8, 3942. [6] S. Miller, G. Fanchini, Y. Y. Lin, C. Li, C. W. Chen, W. F. Su, M. Chhowalla, J. Mater. Chem. 2008, 18, 306. [7] H. Xin, F. S. Kim, S. A. Jenekhe, J. Am. Chem. Soc. 2008, 130, 5424. [8] H. Xin, G. Q. Ren, F. S. Kim, S. A. Jenekhe, Chem. Mater. 2008, 20, 6199. [9] T. A. Bull, L. S. C. Pingree, S. A. Jenekhe, D. S. Ginger, C. K. Luscombe, ACS Nano 2009, 3, 627. [10] P. T. Wu, H. Xin, F. S. Kim, G. Q. Ren, S. A. Jenekhe, Macromolecules 2009, 42, 8817. [11] J. Peet, J. Y. Kim, N. E. Coates, W. L. Ma, D. Moses, A. J. Heeger, G. C. Bazan, Nat. Mater. 2007, 6, 497. [12] N. E. Coates, I. W. Hwang, J. Peet, G. C. Bazan, D. Moses, A. J. Heeger, Appl. Phys. Lett. 2008, 93, 3. [13] J. K. Lee, W. L. Ma, C. J. Brabec, J. Yuen, J. S. Moon, J. Y. Kim, K. Lee, G. C. Bazan, A. J. Heeger, J. Am. Chem. Soc. 2008, 130, 3619. [14] J. C. Bijleveld, A. P. Zoombelt, S. G. J. Mathijssen, M. M. Wienk, M. Turbiez, D. M. de Leeuw, R. A. J. Janssen, J. Am. Chem. Soc. 2009, 131, 16616. [15] J. Peet, N. S. Cho, S. K. Lee, G. C. Bazan, Macromolecules 2008, 41, 8655. [16] F. S. Kim, G. Ren, S. A. Jenekhe, Chem. Mater. 2011, 23, 682. [17] F. Padinger, R. S. Rittberger, N. S. Sariciftci, Adv. Funct. Mater. 2003, 13, 85. [18] G. Li, V. Shrotriya, J. S. Huang, Y. Yao, T. Moriarty, K. Emery, Y. Yang, Nat. Mater. 2005, 4, 864. [19] G. Li, Y. Yao, H. Yang, V. Shrotriya, G. Yang, Y. Yang, Adv. Funct. Mater. 2007, 17, 1636. [20] H. Xin, X. G. Guo, F. S. Kim, G. Q. Ren, M. D. Watson, S. A. Jenekhe, J. Mater. Chem. 2009, 19, 5303. [21] D. Muhlbacher, M. Scharber, M. Morana, Z. G. Zhu, D. Waller, R. Gaudiana, C. Brabec, Adv. Mater. 2006, 18, 2884. [22] A. Pivrikas, H. Neugebauer, N. S. Sariciftci, Sol. Energy 2011, 85, 1226. [23] C. V. Hoven, X.-D. Dang, R. C. Coffin, J. Peet, T.-Q. Nguyen, G. C. Bazan, Adv. Mater. 2010, 22, E63. [24] Y. Liang, Z. Xu, J. Xia, S.-T. Tsai, Y. Wu, G. Li, C. Ray, L. Yu, Adv. Mater. 2010, 22, 1. [25] A. C. Mayer, M. F. Toney, S. R. Scully, J. Rivnay, C. J. Brabec, M. Scharber, M. Koppe, M. Heeney, I. McCulloch, M. D. McGehee, Adv. Funct. Mater. 2009, 19, 1173.

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

581

www.advenergymat.de

FULL PAPER

www.MaterialsViews.com [26] N. C. Cates, R. Gysel, Z. Beiley, C. E. Miller, M. F. Toney, M. Heeney, I. McCulloch, M. D. McGehee, Nano Lett. 2009, 9, 4153. [27] N. C. Cates, R. Gysel, J. E. P. Dahl, A. Sellinger, M. D. McGehee, Chem. Mater. 2010, 22, 3543. [28] S. A. Jenekhe, J. A. Osaheni, Science 1994, 265, 765. [29] S. A. Jenekhe, Adv. Mater. 1995, 7, 309. [30] K. Tvingstedt, K. Vandewal, A. Gadisa, F. L. Zhang, J. Manca, O. Inganas, J. Am. Chem. Soc. 2009, 131, 11819.

582

wileyonlinelibrary.com

[31] X. G. Guo, F. S. Kim, S. A. Jenekhe, M. D. Watson, J. Am. Chem. Soc. 2009, 131, 7206. [32] H. X. Zhou, L. Q. Yang, A. C. Stuart, S. C. Price, S. B. Liu, W. You, Angew. Chem., Int. Ed. 2011, 50, 2995. [33] Y. P. Zou, A. Najari, P. Berrouard, S. Beaupre, B. R. Aich, Y. Tao, M. Leclerc, J. Am. Chem. Soc. 2010, 132, 5330. [34] C. Piliego, T. W. Holcombe, J. D. Douglas, C. H. Woo, P. M. Beaujuge, J. M. J. Frechet, J. Am. Chem. Soc. 2010, 132, 7595.

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Adv. Energy Mater. 2012, 2, 575–582