Polymer Chemistry

0 downloads 0 Views 3MB Size Report
Feb 25, 2015 - Recently, the introduction of alkylthio substituents ... bState Key Laboratory of Polymer Physics and Chemistry, Beijing National Laboratory.
Polymer Chemistry View Article Online

Published on 25 February 2015. Downloaded by Kansas State University on 26/02/2015 06:41:07.

PAPER

Cite this: DOI: 10.1039/c5py00071h

View Journal

Optimization of side chains in alkylthiothiophenesubstituted benzo[1,2-b:4,5-b’]dithiophene-based photovoltaic polymers† Shaoqing Zhang,a,b Mohammad Afsar Uddin,c Wenchao Zhao,b Long Ye,b Han Young Woo,*c Delong Liu,b Bei Yang,b Huifeng Yao,b Yong Cuia,b and Jianhui Hou*a,b Alkyl side chains play critical roles in the molecular design of conjugated polymers for applications in bulk-heterojunction (BHJ) polymer solar cells (PSCs). Recently, the introduction of alkylthio substituents onto poly(benzo[1,2-b:4,5-b’]dithiophene-alt-thieno[3,4-b]thiophene) (PBDTTT)-based conjugated polymers has been proved to be an effective method to improve the photovoltaic properties of the polymers. In this contribution, three alkylthiothiophene-substituted benzodithiophene (BDT-TS) based polymers, named PBDT-TS1, PBDT-TS2 and PBDT-TS3, were synthesized and applied as donor materials in PSCs. In these three polymers, octyl, 2-ethylhexyl and 3,7-dimethyloctyl are used on their BDT units, respectively. The polymers were characterized in parallel by absorption spectroscopy, thermogravimetric analysis (TGA), electrochemical cyclic voltammetry (CV) and grazing-incidence wide-angle X-ray scattering (GI-WAXS), and also their photovoltaic properties in PSCs were studied and compared. The results reveal that the alkyls have little influence on absorption spectra and molecular energy levels of the polymers. The GI-WAXS results show that PBDT-TS1 has stronger and tighter π–π stacking than the other two polymers, implying that linear alkyls may reduce steric hindrance than branched alkyl chains in an aggregation state. As a consequence of the strong π–π inter-chain packing of PBDT-TS1, an increased short circuit current density ( JSC) and fill factor (FF) as well as a power conversion efficiency of over 9.5% are achieved in single-cell BHJ devices, which are obviously higher than those for devices based on the other two

Received 18th January 2015, Accepted 13th February 2015 DOI: 10.1039/c5py00071h www.rsc.org/polymers

polymers. Overall, the results of this work suggest that alkyl side groups play an important role in affecting the π–π stacking of the conjugated polymers, i.e., the linear octyl has weaker steric hindrance for the inter-chain π–π stack than the branched 2-ethylhexyl and 3,7-dimethyloctyl, and for the highly efficient polymer based on the 2-alkylthiothiophene-substituted BDT, PBDT-TS1 has the optimal structure.

Introduction Alkyl side chains play critical roles in the molecular design of conjugated polymers for applications in polymer solar cells (PSCs). As is known, unsubstituted conjugated polymers are commonly insoluble in most organic solvents due to the high rigidity and strong inter-chain π–π interaction of their backa University of Science and Technology Beijing, School of Chemistry and Biology Engineering, Beijing 100083, China b State Key Laboratory of Polymer Physics and Chemistry, Beijing National Laboratory for Molecular Sciences, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China. E-mail: [email protected]; Tel: +86-10-82615900 c Department of Cogno-Mechatronics Engineering, Pusan National University, Miryang 627-706, South Korea † Electronic supplementary information (ESI) available: 1H NMR spectra of the polymers (PBDT-TS1, PBDT-TS2 and PBDT-TS3), the PV certificate report, and 2D patterns of GI-WAXS data of the blend films. See DOI: 10.1039/c5py00071h

This journal is © The Royal Society of Chemistry 2015

bones. When bulky alkyls are introduced as the side groups onto conjugated polymers, the inter-chain π–π interaction can be weakened and also the polymer–solvent interaction can be enhanced, so the solubility of conjugated polymers can be effectively improved. For example, unsubstituted polythiophene is insoluble, while poly(3-hexylthiophene) (P3HT)1 and poly(3-octyl-thiophene) (P3OT)2 can be easily dissolved in many types of solvents such as chloroform, toluene, chlorobenzene etc. Besides solubility, morphological properties of conjugated polymers in the aggregation state can also be modulated by introducing varied alkyls, i.e. the laminar packing and π–π stacking distances, aggregation size, crystallinity and so on, can be tuned by changing the alkyl side chains. As a result, photovoltaic properties of conjugated polymers are often sensitive to the change in the alkyls. For instance, although the chemical structures of P3HT and P3OT are very similar and also show almost identical properties in

Polym. Chem.

View Article Online

Published on 25 February 2015. Downloaded by Kansas State University on 26/02/2015 06:41:07.

Paper

Polymer Chemistry

absorption spectra and molecular energy levels, these two polymers exhibit much different photovoltaic properties.3 Therefore, how to optimize the alkyl side chains is deemed to be an important issue for the molecular design of conjugated polymers for applications not only in PSCs but also in other types of organic electronic devices.4–11 Recently, quite a few interesting studies related to the optimization of alkyls in conjugated polymers have been reported. For example, for polymers based on PFDTBT12 backbones, the polymer substituted by 3,7-dimethyloctyls (BisDMO-PFDTBT)13 showed higher hole mobility than the 2-ethylhexyl-substituted analogue (BisEH-PFDTBT) due to the weaker spacing filling effect of 3,7-dimethyloctyl than 2-ethylhexyl, and thus BisDMO-PFDTBT shows better photovoltaic performance than BisEH-PFDTBT in PSCs. For the polymers based on benzo[1,2-b:4,5-b′]dithiophene (BDT) and thieno-[3,4-c]pyrrole-4,6-dione (TPD), named PBDTTPD,14,15 when the alkyls on their TPD units were changed from octyl to heptyl, the power conversion efficiency (PCE) of the corresponding PSCs can be improved from 7.5% to 8.5%.16 For isoindigo (IID)-based conjugated polymers for applications in organic field effect transistors (OFETs), the crystallinity of the polymer films and thus their mobilities are also very sensitive to the substituting alkyls on their IID units.17 According to the reported study, although the correlations among the substituting alkyls, morphologies and electronic properties of the varied polymers have been well studied, it is still very hard to get a clear regulation to guide the molecular design of conjugated polymers. Therefore, for varied types of conjugated polymers, the alkyls have to be carefully optimized case by case. Very recently, we reported a polymer based on 2-octylthiothienyl-substituted BDT units, named PBDT-TS1, and the PSC

Scheme 1

based on this polymer showed an initial PCE of 9.48%.18 Thereafter, a PCE of 9.87% from internal testing and 10.2% from certification was obtained with the treatment of solvent variation.19 The certified report is provided in ESI (Fig. S1†). Compared to the analogous polymer based on 2-(2-ethylhexylthio)-thienyl-substituted BDT units reported in the previous work by Li et al.,20 it seems that the polymer PBDT-TS1 showed improved photovoltaic properties. However, the correlations among the alkyls, morphological and photovoltaic properties of this type of polymer have not yet been fully investigated. Therefore, it is still of great interest to reveal the influence of alkyl side chains on the photovoltaic properties of this type of polymer. Herein, as shown in Scheme 1, three polymers (PBDT-TS1, PBDT-TS2 and PBDT-TS3) with very similar structures were synthesized, and in these polymers, three types of commonly used alkyls, including octyl, 2-ethylhexyl and 3,7dimethyloctyl, were introduced as the substituents on their BDT units respectively. The polymers were characterized by UV-vis absorption spectroscopy, thermogravimetric analysis (TGA), electrochemical cyclic voltammetry (CV), grazing-incidence wide-angle X-ray scattering (GI-WAXS) etc. The results reveal that the alkyls on the BDT units have little influence on band gaps and molecular energy levels but strong effects on the inter-chain packing properties of the polymer films. As observed in GI-WAXS measurements, PBDT-TS1 shows the strongest and also the tightest π–π stacking in this series of polymers, which may facilitate inter-chain charge diffusion and transport. Moreover, PSCs based on these three polymers were fabricated and characterized in parallel, and the photovoltaic results clearly show that devices based on PBDT-TS1 exhibit a similar open circuit voltage (VOC) but higher short circuit current density ( JSC) and fill factor (FF) than devices

Chemical structures and synthesis routes of PBDT-TS1, PBDT-TS2 and PBDT-TS3.

Polym. Chem.

This journal is © The Royal Society of Chemistry 2015

View Article Online

Polymer Chemistry

Published on 25 February 2015. Downloaded by Kansas State University on 26/02/2015 06:41:07.

based on PBDT-TS2 or PBDT-TS3. Therefore, the PCE of PBDT-TS1-based devices reached 9.5%, which is obviously higher than those of devices based on PBDT-TS2 and PBDT-TS3 respectively. The present work clearly demonstrates the correlations among the alkyls, morphological and photovoltaic properties, and PBDT-TS1 is the best one among this type of polymer.

Results and discussion Synthesis of the polymers As shown in Scheme 1, three types of alkyls, including octyl, 2-ethylhexyl and 3,7-dimethyloctyl, were introduced by using 2-thiophenethiol and alkyl bromides as the starting materials. The 2-alkylthio-thiophenes were added to BDT units to build the so-called two-dimensional conjugated building blocks, and then the trimethyltin functional groups were introduced to produce the BDT-TS monomers as shown in Scheme 1. The three polymers were synthesized using Stille coupling reactions respectively. The corresponding polymers, named PBDT-TS1, PBDT-TS2 and PBDT-TS3, were prepared using the same conditions provided in the Experimental section. The polymers PBDT-TS2 and PBDT-TS3 can be easily dissolved in the commonly used solvents for device fabrication such as chloroform, xylenes, chlorobenzene (CB) and o-dichlorobenzene (DCB), while PBDT-TS1 shows relatively poor solubility in these solvents at room temperature (ca. 20 °C) but good solubility in warm solvents (>45 °C), indicating that PBDT-TS1 may have a stronger inter-chain interaction in the aggregation state than the other two polymers. The molecular weights of the three polymers were determined by gel permeation chromatography (GPC) using trichlorobenzene as the eluent at 150 °C. The resulting number-average molecular weights (Mn) of PBDT-TS1, PBDT-TS2 and PBDT-TS3 are 30.1 k (PDI = 3.3), 16.7 k (PDI = 1.7) and 70.0 k (PDI = 3.12), respectively. The thermal stabilities were evaluated by thermogravimetric analysis (TGA) (see Fig. 1), and the decomposition onset point of PBDT-TS2 is

Paper

at ca. 330 °C, which is about 20 °C lower than the other two polymers, indicating that thermal stability of the polymers can be slightly affected either by the difference of the molecular weight or by the change in the alkyl side groups. Optical and electrochemical properties of the polymers UV-vis absorption spectra of the polymers in a dilute CB solution and solid films are shown in Fig. 2a and 2b, and the detailed absorption data are shown in Table 1. In the solution state, these three polymers display similar absorption spectra in the range from 440 to 790 nm, peaking at ca. 700 nm, and in detail, PBDT-TS1 shows a slightly broader absorption shoulder at the long wavelength direction than the other two polymers. From the solution state to the solid state, the absorption spectra of the polymers are red-shifted, i.e. 16 nm, 13 nm, and 6 nm for PBDT-TS1, PBDT-TS2 and PBDT-TS3, respectively, indicating that PBDT-TS1 may have a stronger inter-chain π–π stacking effect than the other two polymers in the solid state. Overall, all three films show absorption edges at ca. 800 nm, which corresponds to a band gap of 1.55 eV. Obviously, these polymers are ideal absorbers for utilizing sunlight in the visible band. CV measurements were conducted to determine the molecular energy levels of the three polymers by casting thin films on a glassy carbon electrode, and the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) levels were estimated according to the onset potentials of the p-doping and n-doping processes, respectively. Fig. 3 shows the CV plots of these three polymers. It is very clear that the onset potentials of the p-doping processes of PBDT-TS1, PBDT-TS2 and PBDT-TS3 are all at ca. 0.96 V; calculating via EHOMO = −(Eonset − Eonset ox Fc/Fc +) − 4.8 eV, the corresponding HOMO levels of the three polymers are all at −5.29 eV. According to the reported studies, when varied alkyls were used as the substituted side groups, HOMO levels of the conjugated polymers can be little affected.7,10 These results consistently imply that the alkyls used in this family of polymers have little influence on changing the HOMO levels. Since VOC values of bulk heterojunction PSCs are directly proportional to the offset between the HOMO level of the electron donor and the LUMO level of the electron acceptor, PSCs based on the three polymers in this work should possess similar VOC values. GI-WAXS analysis of the polymers

Fig. 1 Thermogravimetric analysis plots of three polymers with a heating rate of 10 °C min−1 under an inert atmosphere.

This journal is © The Royal Society of Chemistry 2015

In this study, GI-WAXS was used to investigate the crystalline packing characteristics of three pristine polymers. Fig. 4(a1– a3) show the 2D profiles of thin films of the polymers, and in Fig. 4b and c, the out-of-plane and in-plane diffraction profiles are provided. As shown in Fig. 4c, a clear (100) peak was observed but higher order (h00) scattering peaks were negligible for all polymers; in detail, the (100) peak of PBDT-TS1 is stronger than that of PBDT-TS3 but weaker than that of PBDT-TS2. In Fig. 4c, the (100) reflection peaks of PBDT-TS1, PBDT-TS2 and PBDT-TS3 were measured at 0.23 Å−1, 0.25 Å−1 and 0.23 Å−1, and the corresponding d-spacings were determined to be 27.3 Å, 25.1 Å and 27.3 Å, respectively. The results

Polym. Chem.

View Article Online

Published on 25 February 2015. Downloaded by Kansas State University on 26/02/2015 06:41:07.

Paper

Fig. 2

Table 1

Polymer Chemistry

UV-vis absorption spectra of three polymers (a) in chlorobenzene and (b) in solid thin films.

Summary of optical and electrochemical properties of the three polymers

PBDT-TS1 PBDT-TS2 PBDT-TS3

Fig. 3

Td (°C)

λmax (nm) soln.

λmax (nm) film

Eopt g (eV)

HOMO (eV)

LUMO (eV)

μhole (cm2 Vs−1)

370 348 367

700 700 700

716 713 706

1.60 1.59 1.61

−5.29 −5.29 −5.29

−3.40 −3.53 −3.49

1.09 × 10−2 1.62 × 10−2 5.01 × 10−3

CV plots of PBDT-TS1, PBDT-TS2, and PBDT-TS3 in thin films.

indicate that the laminar packing distances of the polymers are mainly determined by the length of the alkyl side-chains attached to the BDT core, i.e. the chain length of octyl is similar to that of 3,7-dimethyloctyl but longer than 2-ethylhexyl. The measured inter-lamellar distance shows good agreement with the molecular structures with different alkyl sidechains. In Fig. 4b, all the polymer films show clear π–π stacking (010) reflection peaks in the out-of-plane direction, indicating a face-on orientation. The (010) reflection peak was measured at 1.70 Å−1, 1.64 Å−1 and 1.65 Å−1 for PBDT-TS1, PBDT-TS2 and PBDT-TS3, respectively, showing the corresponding d-spacings of 3.69 Å, 3.83 Å and 3.81 Å, respectively. Obviously, among the three polymers, PBDT-TS1 has the strongest and tightest π–π stacking effect in films, while PBDT-TS2 and PBDT-TS3 show larger d-spacing with weaker π–π stacking.

Polym. Chem.

Moreover, these polymers adopt a face-on orientation in thin films, showing a strong (100) peak in the in-plane direction with a (010) peak in the out-of-plane direction. According to the GI-WAXS results, it can be concluded that the length of the alkyl side-chains on BDT significantly influences the (100) d-spacing, and the branched alkyls (i.e., 2-ethylhexyl and 3,7dimethyloctyl) weaken the π–π inter-chain packing with longer π–π distance. Since the enhanced π–π stacking effects are helpful for improving inter-chain charge transport,21–23 the hole mobilities, μh, of the pristine polymer films were measured by the space-charge-limited current (SCLC) method24 and the mobility data are collected in Table 1. The μh of PBDT-TS1, PBDT-TS2 and PBDT-TS3 are 1.09 × 10−2 cm2 Vs−1, 1.62 × 10−2 cm2 Vs−1 and 5.01 × 10−3 cm2 Vs−1, respectively. Obviously, the morphological properties and thus the higher μh of PBDT-TS1 should be more favorable for achieving better photovoltaic performance in PSCs (Table 2). Photovoltaic properties and thin film morphology PSCs with the device structure of ITO/PEDOT:PSS (30 nm)/ polymer : PC71BM/Mg/Al were fabricated and characterized to investigate the photovoltaic properties of these polymers. The detailed fabrication conditions for making the PSCs based on these three polymers are presented in the Experimental section. In order to obtain the optimal photovoltaic performance in PSCs, the device fabrication processes were optimized in parallel by referring to the reported studies.10,11 Herein, CB was used to prepare solutions of the active layer materials ( polymer : PC71BM), and the concentration of the solutions for the polymers in this work was 10 mg mL−1 ( polymer/CB). Initially, varied D/A ratios ( polymer : PC71BM, wt/wt), 1 : 1, 1 : 1.5 and 1 : 2, were scanned for each of these three polymers. For all three polymers, the optimal D/A ratio is 1 : 1.5, which is

This journal is © The Royal Society of Chemistry 2015

View Article Online

Published on 25 February 2015. Downloaded by Kansas State University on 26/02/2015 06:41:07.

Polymer Chemistry

Fig. 4

Paper

GI-WAXS data of the thin films of PBDT-TS1, PBDT-TS2 and PBDT-TS3: (a) 2D patterns, (b) out-of-plane profiles and (c) in-plane profiles.

Table 2 Packing parameters of pristine polymer films derived from GI-WAXS measurementsa

Crystallographic parameters Pristine polymer films PBDT-TS1 PBDT-TS2 PBDT-TS3 a

Lamellar spacing

π–π stack

Plane

q (Å−1)

d-Spacing (Å)

q (Å−1)

d-Spacing (Å)

In out In out In out

0.233 — 0.251 — 0.231 —

27.3 — 25.02 — 27.18 —

— 1.70 — 1.64 — 1.65

— 3.69 — 3.83 — 3.81

Dash represents no diffraction signal.

consistent with the reported studies for the PBDTTT-series polymers.10,11,25,26 As is known, for the PBDTTT-series polymers, when CB or DCB is used as the processing solvent, the use of 3% 1,8-diiodooctane (DIO) (DIO/CB, v/v) as the solvent additive for the spin-coating process is helpful to form favorable morphologies in the blend films of the polymer and PC71BM and thus can be used to realize better photovoltaic performance in devices. Herein, varied amounts of DIO, 1%, 3% and 5% (v/v, DIO/CB), were scanned to obtain the optimal ratio for the DIO additive, and we found that for all three polymers, the optimal volume ratio of DIO is 3%, which is the same as the reported studies mentioned above. In addition, varied thicknesses of the active layers were scanned by controlling the spin speed for spin-coating, and we found that the optimal active layer thicknesses for devices based on these three polymers are all in the range 90–100 nm. Overall, the

This journal is © The Royal Society of Chemistry 2015

results of optimization of the device fabrications reveal that the alkyls on the BDT units have little influence on the optimal conditions for device fabrication, and PSCs based on this series of polymers can be processed using the same conditions. Since this part of the results is consistent with the reported studies, it is not necessary to make redundant discussions. The photovoltaic properties of the PSCs fabricated under the optimal conditions were characterized. Fig. 5a shows the voltage–current density ( J–V) curves of the PSCs based on these three polymers under illumination (AM1.5G, 100 mW cm−2), and the corresponding photovoltaic parameters are listed in Table 3. According to the reported studies, the conjugated polymers with varied alkyls side-chains often show different VOC values in devices. For example, for the polymer based on BDT and TPD, when the branched alkyls were introduced onto their TPD units, the corresponding PSC device showed higher VOC than the device based on the linear alkylsubstituted analog polymer;9,16 for the polymer based on BDT and TT units, the branched alkyl-substituted polymer also tends to yield higher VOC in the device than the linear alkylsubstituted analogues.18,26 However, the VOC of the devices based on PBDT-TS1, PBDT-TS2 and PBDT-TS3 is 0.807 V, 0.813 V and 0.809 V, respectively, indicating that the alkyls on the BDT units in these three polymers have little influence on VOC. The device based on PBDT-TS2 shows a similar JSC to the device based on PBDT-TS3, while the PBDT-TS1-based device shows a JSC of 17.55 mA cm−2, which is higher than the other two devices. For FF, the PBDT-TS1-based device shows a highest value of 0.672 among these three types of devices. Since the PBDT-TS1-based device shows similar VOC and

Polym. Chem.

View Article Online

Published on 25 February 2015. Downloaded by Kansas State University on 26/02/2015 06:41:07.

Paper

Polymer Chemistry

Fig. 5 (a) Current density–voltage ( J–V) curves under illumination of AM 1.5G (100 mW cm−2); (b) characteristic external quantum efficiency (EQE) spectra of PSCs fabricated from the polymers:PC71BM blends; (c) absorption spectra of the active layers of the devices.

Table 3 Photovoltaic properties of PSCs based on the structure of ITO/ PEDOT:PSS/PBDT-TSx:PC71BM/Mg/Al under illumination of AM 1.5G (100 mW cm−2)

Polymer/ PC71BM

VOC [V]

JSC [mA cm−2]

FF

PCEave [%]

Thickness [nm]

PBDT-TS1 PBDT-TS2 PBDT-TS3

0.807 0.813 0.809

17.55 16.52 16.22

0.672 0.623 0.560

9.52 8.37 7.36

90 102 94

higher JSC and FF among these three types of devices, it shows a PCE of 9.52%, which is higher than the other two types of devices. Moreover, the external quantum efficiencies (EQEs) and absorption spectra of the active layers of these three types of devices are shown in Fig. 5b and c. Interestingly, although the pure polymer films show very similar absorption edges (see Fig. 2b), the polymers : PC71BM blends show different absorptions in the long wavelength region, i.e., the half maximum absorption position at long wavelength of the PBDT-TS1: PC71BM blend film is 782 nm, while those of the blend films

of PBDT-TS2:PC71BM and PBDT-TS3:PC71BM are at 766 and 775 nm (see Fig. 5c), respectively. Obviously, the broader absorption at the long wavelength region should be helpful to obtain better utilization of the sunlight. As a result, the EQE spectrum of the PBDT-TS1-based device is ca. 10 nm broader than those of the PBDT-TS2-based and PBDT-TS3-based devices. In Fig. 5b, it can also be seen that the quantum efficiency of PBDT-TS1 is slightly higher than those of the other two devices in the whole response range. Since the PBDT-TS1-based device shows a higher and broader EQE spectrum among these three types of devices, it is reasonable to realize a higher JSC value in the PBDT-TS1-based device. Morphological properties of the polymer : PC71BM blends Atomic force microscopy (AFM) and transmission electron microscopy (TEM) were employed to gain more information about the morphologies of the polymer/PC71BM blends. Fig. 6 shows the AFM images of the polymer/PC71BM blend films prepared under the optimal conditions for device fabrication. As shown in Fig. 6a–c, the surface morphologies of the blends show barely coarse appearance with the mean square surface

Fig. 6 AFM height images (a–c) and phase images (d–f) of polymer/PC71BM (1 : 1.5 weight ratio) blends cast from chlorobenzene with 3% DIO (v/v) as a processing additive.

Polym. Chem.

This journal is © The Royal Society of Chemistry 2015

View Article Online

Published on 25 February 2015. Downloaded by Kansas State University on 26/02/2015 06:41:07.

Polymer Chemistry

Fig. 7

Paper

TEM images of (a) PBDT-TS1:PC71BM, (b) PBDT-TS2:PC71BM and (c) PBDT-TS3:PC71BM blend thin films.

roughness (Rq) of 1.17 nm, 2.04 nm and 0.70 nm for the films of PBDT-TS1:PC71BM, PBDT-TS2:PC71BM and PBDT-TS3: PC71BM, respectively. The phase images of the three films show very similar morphologies and the aggregation sizes in the blends can be hardly evaluated (see Fig. 6d–f ). The TEM images of the three types of blends are shown in Fig. 7a–c. It seems that all three polymers show good miscibility with PC71BM, and no big size aggregation can be observed in the TEM images. According to the AFM and TEM measurements, it is very hard to get a conclusion for the higher photovoltaic performance of the PBDT-TS1:PC71BM blend. Therefore, the GI-WAXS method was used to provide more information about the blend films. As discussed above, the pristine polymers show weak crystalline structures in GI-WAXS measurements, so after blending with PC71BM, the crystallinities of the films become even weaker. As shown in Fig. S2 (see ESI†), only the reflection of PC71BM and the (100) diffraction peaks of the polymers can be clearly observed in the 2D-patterns of the blend films. For the blend film of PBDT-TS1:PC71BM, a weak (010) reflection can be distinguished, but no (010) reflection can be found for the other two blend films. The results obtained from the GI-WAXS measurements clearly imply that PBDT-TS1 has the most pronounced π–π inter-chain ordering in the blended film than the other two polymers.

Conclusion

PC71BM and PBDT-TS3:PC71BM in AFM and TEM measurements. These three types of polymer : PC71BM blends have amorphous morphologies in GI-WAXS measurements, and only for the blend of PBDT-TS1:PC71BM, a weak π–π stacking packing reflection can be found. The photovoltaic properties of these three polymers were characterized in parallel. We found that the optimal device fabrication conditions of these three polymers are the same. The PSC based on PBDT-TS1 shows a PCE of 9.52%, which is higher than that of the devices based on the other two polymers. Overall, the results of this work suggest that the alkyl side groups play an important role in affecting the inter-lamellar packing and π–π stacking of conjugated polymers, i.e., the linear octyl group has weaker steric hindrance for efficient inter-chain π–π stacking, compared to the branched 2-ethylhexyl and 3,7-dimethyloctyl chains. For the highly efficient polymer based on the 2-alkylthiothiophene-substituted BDT and TT, PBDT-TS1 has the optimal structure. Moreover, this work also suggests that a little change in alkyl side groups may have a great influence on photovoltaic properties, so the optimization of alkyl side groups is an essential step for the molecular design of conjugated polymers for applications in highly efficient PSCs.

Experimental section Materials

Three conjugated polymers, PBDT-TS1, PBDT-TS2 and PBDT-TS3, based on the 2-alkylthiothiophene-substituted BDT and thieno[3,4-b]thiophene (TT) were synthesized and their photovoltaic characteristics were characterized. In these three polymers, octyl, 2-ethylhexyl and 3,7-dimethyloctyl are substituted on the BDT unit, respectively. The results reveal that the alkyl side-chains have little influence on absorption spectra and molecular energy levels of the polymers. The GI-WAXS results show that PBDT-TS1 has the stronger and tighter π–π stacking than the other two polymers, implying that the linear alkyls may have weaker steric hindrance than the branched alkyls in an aggregation state. Although a stronger π–π stacking effect can be observed in the film of PBDT-TS1 than in the other two polymer films, the blend of PBDT-TS1:PC71BM shows a similar film morphology as the blends of PBDT-TS2:

This journal is © The Royal Society of Chemistry 2015

The TT-EF monomer (see Scheme 1) was obtained from Solarmer Materials Inc. Pd(PPh3)4 was purchased from Frontier Chemical Co. Tetrahydrofuran was dried by Na/benzophenone and distilled prior to use. Poly(3,4-ethylene-dioxy-thiophene): poly(styrenesulfonate) (PEDOT:PSS) (Bayer Baytron 4083) was filtered through a 0.45 μm filter prior to use. 2-(Octylthio)thiophene, 2-(2-ethylhexylthio)thiophene and 2-(3,7-dimethyloctylthio)thiophene were synthesized by the reported method.18 The other chemicals, solvents and materials were purchased and used as received. Synthesis Synthesis of 4,8-bis(5-(octylthio)thiophen-2-yl)benzo[1,2b:4,5-b′]dithiophene (1). In a 100 mL two-necked round

Polym. Chem.

View Article Online

Published on 25 February 2015. Downloaded by Kansas State University on 26/02/2015 06:41:07.

Paper

bottom flask, 2-(octylthio)thiophene (4.8 g, 21 mmol) was dissolved in 20 mL of tetrahydrofuran (THF) under an inert atmosphere. Then, 9.3 mL of butyllithium (2.5 M) was added into the flask dropwise at 0 °C using an ice salt bath, and then the reactant was stirred at 50 °C for 20 minutes. Benzo[1,2b:4,5-b′]dithiophene-4,8-dione (1.54 g, 7 mmol) was suspended in 10 mL of THF in another flask with argon protection, and subsequently, the suspension was transferred into the reaction flask in one portion using a syringe. Thereafter, the reactant was stirred at room temperature for 2 hours. Tin dichloride dihydrate (11.07 g, 49 mmol) was dissolved in HCl : H2O (15 mL : 5 mL) and flushed with argon for 10 minutes. After cooling down the reactant to 0 °C, a solution of tin dichloride dihydrate was put into the flask, and the reactant was stirred at room temperature for 6 hours. The reactant was extracted by using petroleum ether and washed with water. After removal of the solvent by rotary evaporation, the raw product was purified by column chromatography ( petroleum ether as an eluent) to afford the title compound as a light yellow solid (3.18 g, yield 70.8%). 1 H NMR (300 MHz, CDCl3), δ( ppm): 7.62–7.61 (d, 2H), 7.49–7.47 (d, 2H), 7.34–7.33 (d, 2H), 7.23–7.22 (d, 2H), 2.95–2.90 (t, 4H), 1.72–1.70 (m, 4H), 1.54–1.28 (m, 20H), 0.90–0.86 (m, 6H). Synthesis of 4,8-bis(5-(2-ethylhexylthio)thiophen-2-yl)benzo[1,2-b:4,5-b′]dithiophene (2). This compound was synthesized by the same procedure as that for compound 1. 1 H NMR (400 MHz, CDCl3), δ( ppm): 7.62–7.61 (d, 2H), 7.48–7.47 (d, 2H), 7.33–7.32 (d, 2H), 7.23–7.22 (d, 2H), 2.97–2.95 (d, 4H), 1.68–1.65 (m, 2H), 1.55–1.33 (m, 16H), 0.95–0.92 (t, 12H). Synthesis of 4,8-bis(5-(3,7-dimethyloctylthio)thiophen-2-yl)benzo[1,2-b:4,5-b′]dithiophene (3). This compound was synthesized by the same procedure as that for compound 1. 1 H NMR (300 MHz, CDCl3), δ( ppm): 7.604 (d, 2H), 7.49–7.47 (d, 2H), 7.34–7.33 (d, 2H), 7.23–7.22 (d, 2H), 2.98–2.90 (m, 4H), 1.56–1.48 (m, 8H), 1.33–1.16 (m, 12H), 0.92–0.90 (d, 6H), 0.87–0.85 (d, 12H). Synthesis of 2,6-bis(trimethylstannyl)-4,8-bis(5-(octylthio)thiophen-2-yl)benzo[1,2-b:4,5-b′]-dithiophene (BDT-TS1). To an argon protected solution of compound 1 (5.17 g, 8.05 mmol) in THF (80 mL), a solution of butyllithium in THF (2.5 M, 7.4 mL) was added dropwise at −78 °C using a liquid nitrogen–acetone bath. After stirring at −78 °C for 30 minutes, trimethyltin chloride (23 mL, 1 M) was added to the reactant and subsequently stirred at room temperature for 20 minutes. The solution was quenched with deionized water, and then extracted by using petroleum ether. The organic layer was separated and concentrated by rotary evaporation, and the crude product was purified by recrystallization via isopropanol– ethanol (1 : 1, v/v). The monomer BDT-TS1 was obtained as a light yellow crystalline solid with a yield of 72%. BDT-TS1: 1H NMR (400 MHz, CDCl3), δ ( ppm): 7.66 (s, 2H), 7.37–7.36 (d, 2H), 7.24–7.23 (d, 2H), 2.95–2.91 (t, 4H), 1.77–1.70 (m, 4H), 1.49–1.42 (m, 4H), 1.32–1.28 (m, 16H), 0.89–0.86 (t, 6H), 0.48–0.34 (t, 18H).

Polym. Chem.

Polymer Chemistry

Calculated for C40H58S6Sn2: C 49.60%, H 6.03%; found: C 49.80%, H 6.02%. Synthesis of 2,6-bis(trimethylstannyl)-4,8-bis(5-(2-ethylhexylthio)thiophen-2-yl)benzo[1,2-b:4,5-b′]-dithiophene (BDT-TS2). This compound was synthesized by the same procedure as that for BDT-TS1. BDT-TS2: 1H NMR (300 MHz, CDCl3), δ ( ppm): 7.64 (s, 2H), 7.34–7.33 (d, 2H), 7.23–7.22 (d, 2H), 2.95–2.93 (d, 4H), 1.68–1.64 (m, 2H), 1.54–1.31(m, 16H), 0.93–0.90 (m, 12H), 0.49–0.40 (t, 18H). Calculated for C40H58S6Sn2: C 49.60%, H 6.03%; found: C 49.74%, H 6.02%. Synthesis of 2,6-bis(trimethylstannyl)-4,8-bis(5-(3.7-dimethyloctylthio)thiophen-2-yl)benzo[1,2-b:4,5-b′]-dithiophene (BDT-TS3). This compound was synthesized by the same procedure as that for BDT-TS1. BDT-TS3: 1H NMR (400 MHz, acetone), δ ( ppm): 7.73 (s, 2H), 7.46–7.45 (d, 2H), 7.34–7.33 (d, 2H), 3.06–2.82 (m, 4H), 1.5 (m, 2H), 1.35–1.31 (m, 18H), 1.20–1.12 (m, 6H), 0.84–0.82 (d, 12H), 0.50–0.34 (t, 18H). Calculated for C44H66S6Sn2: C 51.57%, H 6.49%; found: C 51.81%, H 6.76%. Synthesis of the polymers In a 50 mL round bottom flask, the BDT-TS1, BDT-TS2 or BDT-TS3 monomer (0.53 mmol) and the TT-EF monomer (0.2503 g, 0.53 mmol) were dissolved in a mixed solvent of toluene (20 mL) and DMF (4 mL). After flushing with argon for 5 minutes, 27 mg of Pd(PPh3)4 was added into the flask, and then the mixture was flushed with argon for another 20 minutes and stirred at 110 °C for 16 hours under an argon atmosphere. Subsequently, the solution was cooled down to room temperature and precipitated into 100 mL of methanol. The polymer was collected by filtration and subjected to Soxhlet extraction with acetone (12 hours), hexane (12 hours) and 150 mL of chloroform successively. After concentrating the chloroform solution via rotary evaporation, the polymer was precipitated into 100 mL of methanol. The polymer was collected by filtration with a yield of 50–60%. PBDT-TS1: Calculated for C49H57FO2S8: C 61.72%, H 6.03%; found: C 61.91%, H 6.09%. PBDT-TS2: Calculated for C49H57FO2S8: C 61.72%, H 6.03%; found: C 62.00%, H 6.09%. PBDT-TS3: Calculated for C53H65FO2S8: C 63.05%, H 6.49%; found: C 62.74%, H 6.51%. The 1H NMR data of the three polymers are provided in ESI (Fig. S3†). Fabrication of a regular structure single cell Patterned indium tin oxide (ITO)-coated glass with a sheet resistance of 10–15 ohm per square was cleaned by using a surfactant scrub in acetone and then washed with deionized water, acetone and isopropanol, successively. After UV-ozone cleaning for 15 min, a thin layer (∼30 nm) of PEDOT:PSS (Bayer Baytron 4083) was spin coated onto the ITO substrate and then dried in an oven at 150 °C for 15 min. The substrates were transferred into a nitrogen-filled glove box (