A low cost and high performance polymer donor material for polymer solar cells — TEI-tagged view

Chenkai SunpersName, Fei PanpersName, Haijun BinpersName, Jianqi ZhangpersName, Lingwei XuepersName, Beibei QiupersName, Zhixiang WeipersName, Zhi-Guo ZhangpersName, Yongfang LipersName (2018date). “A low cost and high performance polymer donor material for polymer solar cells.” Nature Communicationstitle·j.

Object of study — the polymer donorterm·key PTQ10term in polymer solar cellsterm·dom with the acceptor IDICterm

Research question — developing low-cost and efficient polymer donorsterm·key for polymer solar cellsterm·dom

Contribution — PTQ10term as a promising polymer donorterm·key for commercial application of polymer solar cellsterm·dom

Full text — tagged

A low costterm·key and high performance polymer donorterm·key material for polymer solar cellsterm·dom Abstract The application of polymer solar cellsterm·dom requires the realization of high efficiency, high stability, and low costterm·key devices. Here we demonstrate a low-cost polymer donorterm·key poly[(thiophene)-alt-(6,7-difluoro-2-(2-hexyldecyloxy)quinoxaline)] (PTQ10term), which is synthesized with high overall yield of 87.4% via only two-step reactions from cheap raw materials. More importantly, an impressive efficiency of 12.70% is obtained for the devices with PTQ10term as donor, and the efficiency of the inverted structured PTQ10-based device also reaches 12.13% (certificated to be 12.0%). Furthermore, the as-cast devices also demonstrate a high efficiency of 10.41% and the devices exhibit insensitivity of active layer thicknessterm·key from 100 nm to 300 nm, which is conductive to the large area fabrication of the devices. In considering the advantages of low costterm·key and high efficiency with thickness insensitivity, we believe that PTQ10term will be a promising polymer donorterm·key for commercial application of polymer solar cellsterm·dom.

Introduction Polymer solar cellsterm·dom (PSCsterm·dom) have received widespread interests and have developed quickly in recent years because of its advantages of solution processing, light weight and flexibility in comparison with the traditional silicon-based solar cells 1 , 2 . Active layerterm·dom of the PSCsterm·dom is composed of a p -type conjugated polymer as donor blending with a fullerene derivative or a nonfullerene n -type organic semiconductor ( n -OS) as acceptor 3 – 5 . Photovoltaic power conversion efficiencyterm·dom (PCEterm·dom), stability, and cost are the three most crucial issues that must be taken into account for the commercial application of PSCsterm·dom. And the donor and acceptor photovoltaic materials play essential role for increasing PCEterm·dom, improving stability and decreasing cost of the PSCsterm·dom.

The concept of bulk-heterojunctionterm·dom PSCsterm·dom was first proposed in 1995 (ref. 3 ), and PCEterm·dom of the PSCsterm·dom at that time was only ca. 1% which is far from considering application. Since then, therefore, the researchers mainly focused on increasing PCEterm·dom of the PSCsterm·dom by developing efficient photovoltaic materials (donor 6 – 8 and acceptor 9 , 10 ) and efficient electrode buffer layer materials 11 – 16 , and optimizing active layerterm·dom morphology 17 – 19 , etc. For the PSCsterm·dom with fullerene derivatives (especially PC 71 BM) as acceptors, PCEterm·dom has gradually increased to over 10% recently by designing low-bandgap polymer donorsterm·key such as PTB7-Th 20 , PNTz4T 21 , and PffBT4T-C 9 C 13 (ref. 22 ), etc. In recent 2 years, low-bandgap nonfullerene n -OS acceptors, such as IDTBR 23 , m -ITIC 24 , IDICterm 25 , etc., have attracted great attention owing to their advantages of broad and strong absorption, easy tuning electronic energy levels, and high morphology stability in comparison with the fullerene derivative acceptors 26 . PCEterm·dom of the PSCsterm·dom with n -OS as acceptors has rapidly increased to over 11% by using medium bandgap polymer donorsterm·key such as J71 (ref. 27 ), PBDB-T 28 , PB3T 29 , etc.

Now, the PCEterm·dom has reached the threshold for application. Next step, we should consider the stability and cost issues for commercial application of the PSCsterm·dom. However, very few efforts have been made on reducing the costs of the photovoltaic materials, and costs of the efficient polymer donorsterm·key reported so far were too high to meet commercial application of the PSCsterm·dom due to their complicated molecular structures, verbose multi-steps synthesis, and multiple purifications 30 . Actually, poly(3-hexylthiophene)term (P3HTterm) is still the main donor material for the fabrication of large area PSCsterm·dom 31 , because it can be synthesized in large scale with relatively low costterm·key. However, the photovoltaic performance of P3HTterm is poor 32 , 33 . Hence, developing low-cost and efficient polymer donorsterm·key becomes one of the greatest challenges for the application of PSCsterm·dom.

Herein, we design and synthesize a low-cost polymer donorterm·key poly [(thiophene)-alt-(6,7-difluoro-2-(2-hexyldecyloxy)quinoxaline)] (PTQ10term) (Fig. 1a ). The molecular design strategy of PTQ10term is based on the donor–acceptorterm·dom (D–Aterm·dom) copolymerization concept, using simple thiophene ring as donor unit and difluorine-substituted quinoxaline as acceptor unit. The alkoxy side chain on quinoxaline unit is to ensure good solubility and to enhance absorption of the polymer, while the difluorine substituentsterm·dom are for down-shifting the highest occupied molecular orbitalterm·dom (HOMOterm·dom) energy level and increasing hole mobilityterm·dom of the polymer donorterm·key 34 . PTQ10term possesses a simple molecular structure and can be synthesized with low costterm·key and high overall yield of 87.4% via only two-step reactions from cheap raw materials. More importantly, the optimized PSCsterm·dom with PTQ10term as donor and an n -OS IDICterm as acceptor demonstrate an impressive PCEterm·dom of 12.70% which is one of the highest PCEterm·dom among the single-junction PSCsterm·dom reported so far, and the as-cast devices without any post-processing also demonstrate a high PCEterm·dom of 10.41%. Furthermore, the devices have good reproducibility and have high tolerance of the active layer thicknessterm·key with a PCEterm·dom over 10% even at an active layer thicknessterm·key of 310 nm. The results indicate that PTQ10term is a promising polymer donorterm·key for commercial products, and it will make the application of PSCsterm·dom highly promising.

Photovoltaic materials and device structure of the PSCsterm·dom. a Molecular structures of the polymer donorterm·key PTQ10term and the n -OS acceptor IDICterm. b Devices architecture of the traditional structured PSCsterm·dom. c Synthetic route of PTQ10term. d Energy level diagram of the related materials used in the PSCsterm·dom. e Normalized absorption spectra of the donor PTQ10term and the acceptor IDICterm Results Synthesis and characterization of PTQ10term The synthetic route of PTQ10term is depicted in Fig. 1c , and the detailed synthesis procedures are described in the Methods section. The synthetic route of monomer 2 was carefully designed with cheap raw material and efficient reaction for realizing low-cost synthesis. PTQ10term possesses good solubility in common organic solvents. Its thermal decomposition temperature ( T d ) at 5% weight loss is measured to be 383 °C (Supplementary Fig. 1a ), indicating its good thermal stability for the application in PSCsterm·dom.

Electronic energy levels of PTQ10term were measured by electrochemical cyclic voltammetry. The E HOMOterm·dom and E LUMO of PTQ10term were calculated to be −5.54 eV and −2.98 eV (Fig. 1d ) from onset oxidation and onset reduction potentials, respectively (Supplementary Fig. 1b ). Figure 1e shows UV–vis absorption spectra of PTQ10term in chloroform solution and in thin film, and the absorption spectrum of IDICterm film for comparison. PTQ10term film displays a strong absorption from 450 to 620 nm with an absorption edge at 645 nm which corresponds to an optical bandgap of 1.92 eV. PTQ10term and IDICterm films display complementary absorption in the wavelength region from 400 to 800 nm, which will benefit to the solar light harvest for the PSCsterm·dom with PTQ10term as donor and IDICterm as acceptor.

Photovoltaic properties In order to investigate photovoltaic properties of PTQ10term, we fabricated the traditional structured PSCsterm·dom with PTQ10term as donor, n -OS IDICterm as acceptor, PEDOT: PSS (poly(3,4-ethylenedioxythiophene):poly(styrene-sulfonate) as anode buffer layer and PDINO (perylene diimide functionalized with amino N-oxide) as cathode buffer layer 16 (Fig. 1b ). It should be mentioned that IDICterm 25 , 35 was selected as acceptor because it possesses a simpler structure with alkyl side chains on its smaller fused ring core and relatively low-cost synthesis in comparison with the widely used n -OS acceptors ITICterm 10 , etc. Photovoltaic performances of the PSCsterm·dom were optimized by using different donor/acceptor weight ratio and different active layer thicknessterm·key, and by the treatment of thermal annealingterm·mth (TAterm·mth) and solvent vapor annealingterm·mth (SAterm·mth). The optimized device fabrication conditions include the donor/acceptor weight ratio of 1:1, active layer thicknessterm·key of 130 nm, TAterm·mth at 140 °C for 5 min, and SAterm·mth by chloroform solvent for 30 s. Figure 2a shows current density–voltage ( J – V ) curves of the optimal PSCsterm·dom based on PTQ10term: IDICterm with or without TAterm·mth and SAterm·mth treatments, and the corresponding photovoltaic parameters are listed in Table 1 .

Photovoltaic parameters of the PSCsterm·dom based on PTQ10term: IDICterm Photovoltaic performance of the PSCsterm·dom based on PTQ10term: IDICterm. a J – V curves of the traditional structured PSCsterm·dom based on PTQ10term: IDICterm (1:1, w/w), under the illumination of AM1.5G, 100 mW cm −2 . b EQEterm·dom spectra of the corresponding PSCsterm·dom. The dependence of J sc ( c ) and V oc ( d ) on light intensity ( P light ) of the optimized PSCsterm·dom It can be seen from Table 1 that all of the PSCsterm·dom exhibit high V oc of 0.960~0.995 V, which should be benefitted from the lower-lying E HOMOterm·dom (−5.54 eV) of PTQ10term. The V oc decreased from 0.995 V to 0.972 V and 0.969 V when the devices were treated with TAterm·mth and TAterm·mth + SAterm·mth, respectively, which could be ascribed to the red-shifted absorption (the reduced optical band gap) of PTQ10term and IDICterm treated with TAterm·mth and TAterm·mth + SAterm·mth (Supplementary Fig. 2a and 2b ). The as-cast PSCsterm·dom without post-processing show an impressive PCEterm·dom of 10.41%, and the PCEterm·dom increased to 11.65% and 12.70%, respectively, after TAterm·mth and TAterm·mth + SAterm·mth treatments. To our knowledge, the PCEterm·dom of 12.70% is one of the highest efficiencies among the single-junction PSCsterm·dom reported to date. In addition, it is worth noticing that PCEterm·dom of 10.41% for the as-cast devices is the highest efficiency in the PSCsterm·dom without post-treatments, and the simple device fabrication process for the as-cast devices will significantly reduce device fabrication costs, which is very important for future industrial production of the PSCsterm·dom.

The external quantum efficiencyterm·dom (EQEterm·dom) spectra of the optimized devices with different post-processing treatments are shown in Fig. 2b . All the three PSCsterm·dom exhibit broad light response with high EQEterm·dom values over 50% from 450 to 730 nm, which means high photoelectric conversion efficiency in the PTQ10term: IDICterm blend films. The current density values integrated from the EQEterm·dom spectra under the AM 1.5G spectrum are 15.30 mA cm −2 for the as-cast device, 16.01 mA cm −2 for the TA-treated device, and 17.08 mA cm −2 for the TAterm·mth + SA-treated device, which are consistent quite well with the J sc values obtained from J – V curves within 4% mismatch, indicating the reliability of the measured J sc data. The enhanced EQEterm·dom value and current density of the PSCsterm·dom with the TAterm·mth or TAterm·mth + SAterm·mth treatments should be ascribed to its red-shifted absorption, enhanced absorption coefficient (Supplementary Fig. 2c ), and the broadened EQEterm·dom spectra (Fig. 2b ) in comparison with that of the blend films without the treatment (as-cast).

In order to confirm the high PCEterm·dom of the PTQ10-based PSCsterm·dom, we fabricated the inverted structured PSCsterm·dom with the device structure of ITO/ZnO/PTQ10term: IDICterm/MoO 3 /Ag, in considering the better stability of the inverted PSCsterm·dom for sending the devices out to the National Institute of Metrologyterm (NIMterm) of China for the efficiency confirmation. The inverted PSCterm·dom based on PTQ10term: IDICterm (1:1.5, w/w) with active layer thicknessterm·key of 130 nm and with the treatment of TAterm·mth at 140 °C for 5 min and SAterm·mth by chloroform solvent for 30 s showed a PCEterm·dom of 12.13% with a V oc of 0.960 V, a J sc of 19.65 mA cm −2 , and an fill factorterm·dom (FFterm·dom) of 64.29%, as shown in Table 1 and in Supplementary Fig. 3 . The PCEterm·dom of the inverted PSCsterm·dom was confirmed to be 12.0% by NIMterm (see the Test Report of NIMterm in Supplementary Fig. 4 and the last line in Table 1 ). The slightly lower PCEterm·dom of the inverted device is due to its lower fill factorterm·dom which could be ascribed to the un-optimized cathode and anode buffer layer materials of the inverted PSCsterm·dom.

Batch to batch variation of the polymers is an unfavorable factor to the commercial application of PSCsterm·dom 36 . Supplementary Table 1 lists the photovoltaic parameters of the optimized traditional structured PSCsterm·dom based on PTQ10term: IDICterm with using the PTQ10term samples synthesized in five batches to investigate the photovoltaic repeatability of PTQ10term. The photovoltaic performance of PTQ10term show less batch to batch variation with the PCEterm·dom values ranging from 11.90% to 12.70%, which indicates the good photovoltaic repeatability of PTQ10term.

The hole mobilityterm·dom ( μ h ) and electron mobility ( μ e ) of the PTQ10term: IDICterm blend layers without (as-cast) and with the TAterm·mth and TAterm·mth + SAterm·mth treatments were measured using space charge limited currentterm·mth (SCLCterm·mth) method with hole-only (ITO/PEDOT: PSS /PTQ10term: IDICterm/Au) and electron-only (ITO/ZnO/PTQ10term: IDICterm/PDINO/Al) devices, and the measurement results are shown in Supplementary Fig. 5 . For the as-cast PTQ10term: IDICterm blend films, μ h and μ e are 0.36 × 10 –4 cm 2 V −1 s −1 and 3.43 × 10 –4 cm 2 V −1 s −1 , respectively, with μ e / μ h of 9.53. While μ h and μ e increased to 3.21 × 10 –4 cm 2 V −1 s −1 and 4.80 × 10 –4 cm 2 V −1 s −1 , respectively, after TAterm·mth treatment with a balanced μ e / μ h of 1.50. With TAterm·mth + SAterm·mth treatment, the μ h and μ e values were further improved to 5.04 × 10 –4 cm 2 V −1 s −1 and 6.72 × 10 –4 cm 2 V −1 s −1 , respectively, with more balanced μ e / μ h ratio of 1.33. The increased and more balanced charge mobility suggests better charge transfer capability of the PTQ10term: IDICterm blend films after TAterm·mth or TAterm·mth + SAterm·mth treatments. This finding together with the enhanced absorption and the reduced charge carrier recombination after TAterm·mth + SAterm·mth treatment could be collectively responsible for the improved J sc and FFterm·dom of the optimized devices.

For application of the PSCsterm·dom, stability of the devices is one of the crucial issues besides photovoltaic performance and cost 37 . Here, the device stability was tested for the inverted PSCsterm·dom based on PTQ10term: IDICterm with simple encapsulation by ultraviolet-curable epoxy and thin glass slides. Supplementary Fig. 6 shows the results of devices stability experimentsterm·mth. The PCEterm·dom of the inverted-structured PSCsterm·dom based on PTQ10term: IDICterm remained 88.27% and 87.82% of their initial value after approximately 1000 h of storage under N 2 and air atmosphere, respectively. Then the efficiency remain almost unchanged in the following 1000 h. The results indicate good stability of the PSCsterm·dom based on PTQ10term: IDICterm.

Charge carrier recombination To understand the effect of TAterm·mth or TAterm·mth + SAterm·mth treatments on the enhanced photovoltaic performance, the charge carrier recombination behavior in the traditional structured PSCsterm·dom was studied by measuring the dependence of J sc and V oc on light intensity ( P light ). The relationship of J sc and P light can be depicted by the formula J sc ∝( P light ) α , where the value of α indicates the degree of bimolecular recombinationterm·dom. The value of α should be 1 when bimolecular recombinationterm·dom do not occur in donor/acceptor blend films, and there is some bimolecular recombinationterm·dom if α value is smaller than 1 (ref. 38 ). Figure 2c displays the plots of log J sc versus log P light , and α values are 0.95, 0.96, and 1.00 for the devices without (as-cast), with TAterm·mth, and with TAterm·mth + SAterm·mth treatments, respectively. The gradually increased values of α indicate the reduced bimolecular recombinationterm·dom when the blend films are processed with TAterm·mth and TAterm·mth + SAterm·mth compared to the as-cast devices. Especially, α value of 1 for the PSCsterm·dom with TAterm·mth + SAterm·mth treatment indicates that there is no bimolecular recombinationterm·dom in the TAterm·mth + SAterm·mth treated devices. For the as-cast devices and TA-treated devices, another plausible reason for the deviation of the α values from unity, can be understood in term of the build-up of space-charge in the device due to the unbalanced electron-hole mobility as indicated by Blom’s work 39 , 40 . Figure 2d shows the plots of V oc versus ln ( P light ) of the PSCsterm·dom. If bimolecular recombinationterm·dom is the exclusive recombination form, the slope of the fitting straight line of V oc versus ln ( P light ) should be kT / e (where e is the elementary charge, k is the Boltzmann constant, and T is the Kelvin temperature) 41 . The slopes of the fitting lines for the as-cast, TA-treated, and TAterm·mth + SA-treated devices are 0.920 kT/e , 0.924 kT/e , and 0.988 kT/e , respectively. The slope very close to kT/e for the TAterm·mth + SA-treated PSCsterm·dom indicates that almost no other recombination occurs in the devices with TAterm·mth + SAterm·mth treatment. The results of J sc and V oc dependence on P light indicate that there are very little charge carrier recombination in the optimized PSCsterm·dom treated with TAterm·mth + SAterm·mth, which consequently results in the best PCEterm·dom of 12.70% for the PSCsterm·dom with TAterm·mth + SAterm·mth treatment.

Morphological characterization Morphology of the active layerterm·dom is a critical factor to determine the photovoltaic performance of the PSCsterm·dom 42 , 43 . Here, the grazing incident wide-angle X-ray diffractionterm·mth (GIWAXSterm·mth) was employed to study the effect of different post treatments on the molecular packing and material crystallinity features within the PTQ10term: IDICterm blend films. Figure 3 shows the plots and images of GIWAXSterm·mth measurements. For the neat PTQ10term film, the laminar diffraction peaks and π–π stacking diffraction peaks located at 0.28 Å −1 and 1.76 Å −1 (Fig. 3a, f ) respectively, corresponding to the lamellar distance of 22.44 Å and π–π stacking distance of 3.57 Å, and the neat IDICterm film shows the lamellar distance of 15.71 Å and π–π stacking distance of 3.50 Å (Fig. 3b, g ). The strong π–π stacking diffraction peaks in the out-of-plane (OOP) direction and weak π–π stacking diffraction peaks in the in-plane (IP) direction of both neat PTQ10term and IDICterm film suggest strong preference of face-on orientation in the vertical direction of substrate for the molecular packing, which is beneficial for efficient charge transport. The GIWAXSterm·mth plots of blend films demonstrate microstructural features of its individual components. For the PTQ10term: IDICterm blend films with TAterm·mth treatment (Fig. 3d, i ), the molecular packing exhibit preferred and enhanced face-on orientation at 1.81 Å −1 with stronger and sharper π–π stacking peaks in OOP direction in comparison with the as-cast films. Moreover, the π–π stacking diffraction peaks intensity at 1.81 Å −1 in OOP direction was further improved with TAterm·mth + SAterm·mth treatment (Fig. 3e, j ), indicating the enhanced charge transport behavior in the vertical direction of substrate in the devices treated by TAterm·mth or TAterm·mth + SAterm·mth. Furthermore, the π–π stacking distance is decreased to 3.47 Å for the TA-treated and TAterm·mth + SA-treated blend films in comparison with the as-cast films with the π–π stacking distance of 3.53 Å, suggesting more tighter molecular packing after the treatment. The results indicate that the preferred face-on orientation, the closer π–π stacking, and the higher crystalline characteristics of the post-treated blend films (especially for the TAterm·mth + SA-treated blend films), assisted charge transport, suppressed charge carrier recombination, and eventually improved the photovoltaic performance. Furthermore, transmission electron microscope (TEM) measurements were carried out to study the effect of processing conditions on the morphology. From the TEM images (Supplementary Fig. 7 ), the PTQ10term: IDICterm blend films show obviously fibrillary networks and increased domain size after TAterm·mth or TAterm·mth + SAterm·mth treatments.

Plots and images of the GIWAXSterm·mth measurements. Line cuts of the GIWAXSterm·mth images of neat PTQ10term film ( a ), neat IDICterm film ( b ), and PTQ10term: IDICterm blend films without (as-cast) ( c ), with TAterm·mth treatment ( d ), and with TAterm·mth + SAterm·mth treatment ( e ). GIWAXSterm·mth images of neat PTQ10term film ( f ), neat IDICterm film ( g ), and PTQ10term: IDICterm blend films without (as-cast) ( h ), with TAterm·mth treatment ( i ), and with TAterm·mth + SAterm·mth treatment ( j ) In addition, photoinduced force microscopyterm·mth (PiFMterm·mth) 44 , an emergent technology that demonstrates the spatially nm-scale patterns of the individual chemical components in their blend films 45 , was used to study the effect of processing conditions on the morphology. The PiFMterm·mth images at the characteristic Fourier transform infrared (FTIR) wavelengths corresponding to absorption peaks of polymer donorterm·key PTQ10term (805 cm −1 ) and acceptor IDICterm (1703 cm −1 ) with different post treatments are shown in Fig. 4 . From the PiFMterm·mth images, the TA-treated PTQ10term: IDICterm blend films (Fig. 4b ) show obviously increased phase separation and domain size in comparison with the as-cast films (Fig. 4a ), and the further increased phase separation and domain size was observed in the TAterm·mth + SA-treated films (Fig. 4c ). The PiFMterm·mth results are consistent with the GIWAXterm·mth and TEM measurements. The results indicate that the gradually enhanced photovoltaic properties of the devices with TAterm·mth and TAterm·mth + SAterm·mth treatment could be ascribed to the larger phase domains and the more continuous donor/acceptor nano-scale phase-separated interpenetrating networks.

FTIR spectra and PiFMterm·mth topography images. FTIR spectra and PiFMterm·mth images of PTQ10term: IDICterm blend films based on FTIR absorption at different wave numbers (PTQ10term, 805 cm −1 and IDICterm, 1703 cm −1 ): without (as-cast) ( a ), with TAterm·mth treatment ( b ), and with TAterm·mth + SAterm·mth treatment ( c ) Thickness dependence of the photovoltaic performance For large area fabrication of the PSCsterm·dom, the active layer thicknessterm·key is difficult to be precisely controlled 46 . Therefore, it is crucial to develop the thickness-insensitive polymer donorsterm·key with excellent photovoltaic performance. Here, we investigated the effect of active layer thicknessterm·key on the photovoltaic performance of the traditional structured PSCsterm·dom based on PTQ10term: IDICterm with active layer thicknessterm·key ranging from 60 to 310 nm. Figure 5a, b shows the thickness dependence of photovoltaic performance, and Supplementary Table 2 lists the corresponding photovoltaic parameters of the devices. The V oc values are nearly constant with a slightly decrease for the active layersterm·dom thicker than 130 nm (Fig. 5a ). The J sc values show an increasing trend from approximately 15 to 19 mA cm −2 with the increase of the active layer thicknessterm·key. The changes of J sc should be the trade-off results between absorbance and charge recombinationterm·dom, the thicker PTQ10term: IDICterm active layersterm·dom will enhance the light harvest which is beneficial to higher J sc , but it also increases charge recombinationterm·dom which will decrease J sc . FFterm·dom shows relatively significant thickness-dependent behavior, the FFterm·dom values remain high and close to 72% even with the active layer thicknessterm·key of up to 210 nm, but it sharply decreased to ca. 58% as the active layer thicknessterm·key increased to 310 nm (Fig. 5b ), which could be due to the increased series resistance of the PSCsterm·dom with the too thick active layerterm·dom. As a result, the highest PCEterm·dom of 12.70% is obtained for the device with the active layer thicknessterm·key of 130 nm. It should be noted that the high PCEterm·dom of 11.59% was also obtained even the active layer thicknessterm·key increased to 210 nm. Amazingly, even for the PSCsterm·dom with a thicker active layerterm·dom of 310 nm, its PCEterm·dom still reached a high value of 10.31%. The excellent and thickness-insensitive photovoltaic performance of the PSCsterm·dom based on PTQ10term: IDICterm makes it a strong candidate for large area fabrication and commercial applications of the PSCsterm·dom.

Thickness dependence of the photovoltaic performance. Plots of V oc or J sc ( a ) and FFterm·dom or PCEterm·dom ( b ) vs. the active layer thicknessterm·key ranging from 60 to 310 nm for the traditional structured PSCsterm·dom Cost and PCEterm·dom analysis PTQ10term has extremely simple molecular structure (Fig. 1a ) in comparison with the efficient polymer donorsterm·key reported in literatures, and it can be easily synthesized via only two-step reactions from initial raw materials and once purification. Besides, all of the raw materials, such as 3,6-dibromo-4,5-difluorobenzene-1,2-diamine, glyoxylic acid, 1-bromo-2-hexyldecane, and 2,5-bis(trimethylstannyl)thiophene, are inexpensive and available from bulk chemical suppliers. Consequently, PTQ10term exhibits extremely high synthetic accessibility and low costterm·key for commercial production, potentially reducing the energy pay back times.

Figure 6a, b displays the plots of PCEterm·dom values versus synthesis steps and overall yield of PTQ10term respectively in comparison with the efficient polymer donorsterm·key with PCEterm·dom over 10% reported in literatures. The corresponding statistical photovoltaic parameters are listed in Supplementary Table 3 . It can be seen from the figures that the PTQ10-based PSCsterm·dom has the highest PCEterm·dom (12.70%) with the minimum synthesis steps of 2 steps (which is ca. one-third or one-fifth of that for the other efficient polymer donorsterm·key) and the highest overall yield of 87.4% (which is ca. 5–20 times of that for the other efficient polymer donorsterm·key). The less synthetic steps of PTQ10term should be ascribed to its simplest D–Aterm·dom structure, and the high overall yield is benefitted from the high yield of its two stepwise reactions (91% and 96%) and only once purification. The ultimate cost of organic photovoltaic materials reduces linearly with the reduction of the number of synthetic steps 30 , thus the ultimate cost of PTQ10term is only few tenths of other efficient donors. Besides, the high overall yield further increases its low-cost advantages. Obviously, PTQ10term possesses great superiority in both cost and photovoltaic performance, which will lead to a bright future for the commercial application of PSCsterm·dom.

Cost and PCEterm·dom analysis of the PSCsterm·dom. Plots of PCEterm·dom vs. synthesis steps ( a ) and overall yield ( b ) of the polymer donorsterm·key reported in literatures with PCEterm·dom over 10% Discussion In conclusion, a low-cost polymer donorterm·key PTQ10term with only two synthetic steps and high yield of 87.4% was designed and synthesized in pursuing low-cost polymer donorterm·key materials for future application of PSCsterm·dom. PTQ10term possesses a broad and strong absorption band in the wavelength range of 450–620 nm with a medium bandgap of 1.92 eV and lower lying HOMOterm·dom energy level of −5.54 eV. The traditional structured PSCsterm·dom using PTQ10term as donor and a narrow bandgap n -OS IDICterm as acceptor demonstrated a high PCEterm·dom of 12.70%, and its photovoltaic performance exhibits insensitivity of active layer thicknessterm·key between 100 and 300 nm, which is conductive to the large area fabrication of PSCsterm·dom. The PCEterm·dom of the inverted structured PSCsterm·dom based on PTQ10term: IDICterm also reached 12.13% which was confirmed to be 12.0% by NIMterm. In comparison with the polymer donorsterm·key reported in literatures with PCEterm·dom over 10%, PTQ10term shows the great advantages of low costterm·key (benefiting by less synthetic steps and high overall yield) and high photovoltaic performance. Therefore, we believe that PTQ10term is a highly promising polymer donorterm·key for large area fabrication of PSCsterm·dom, and it will push the commercial application of PSCsterm·dom forward.

Methods Materials and synthesis IDICterm was purchased from Solarmer Materials Inc. Other chemicals and solvents were obtained from J&K, Alfa Aesar, and TCI Chemical Co., respectively. The monomer compound 2 and polymer PTQ10term was synthesized according to the synthetic route shown in Fig. 1c . The detailed synthesis procedures are described in the following 47 .

5,8-dibromo-6,7-difluoro-2-(2-hexyldecyloxy)quinoxaline (compound 2): To a two-necked, round-bottom flask, 3,6-dibromo-4,5-difluorobenzene-1,2-diamine (compound 1) (906 mg, 3 mmol), glyoxylic acid (222 mg, 3 mmol), and acetic acid (30 mL) are added. The mixture is warmed to 40 °C for 10 min, and then the solution is stirred at room temperature for 3 h. The precipitate is collected by filtration and dried to get a white solid without further purification. The white solid (1 g, 2.94 mmol), potassium tert-butanolate (395 mg, 3.53 mmol), and 1-bromo-2-hexyldecane (897 mg, 2.94 mmol) are dissolved in methanol (30 mL). The mixture is refluxed for 12 h, then cooled to room temperature. After that, the reaction mixture is poured into saturated NH 4 Cl solution, extracted with dichloromethane, and washed with water. The organic extraction is dried over anhydrous MgSO 4 , and the solvent is evaporated under reduced pressure. Compound 2 is obtained as colorless oil (1.54 g, 2.73 mmol) from the product through column chromatography on silica gel, with an overall yield of 91%. 1 H NMR (400 MHz, CDCl 3 ): δ (p.p.m.) 8.50 (s, 1H), 4.49 (d, J = 5.7 Hz, 2H), 1.95–1.86 (m, 1H), 1.55–1.36 (m, 8H), 1.35–1.18 (m, 16H), 0.92–0.83 (m, 6H). 13 C NMR (100 MHz, CDCl 3 ): δ (p.p.m.) 158.69, 151.82, 149.33, 146.88, 140.66, 136.31, 133.26, 109.75, 107.60, 70.50, 37.46, 31.87, 31.36, 29.98, 29.61, 29.31, 26.84, 22.67, 14.09.

Poly[(thiophene)-alt-(6,7-difluoro-2-(2-hexyldecyloxy)quinoxaline] (PTQ10term): The polymer PTQ10term is synthesized according to still-coupling poly-condensation between compound 2 and 2,5-bis(trimethylstannyl)thiophene under protection of argon. Compound 2 (112.8 mg, 0.2 mmol), 2,5-bis(trimethylstannyl)thiophene (82 mg, 0.2 mmol), and anhydrous toluene (10 mL) are added to a 25-mL double-neck round-bottom flask. The flask is flushed with argon for 10 min, and then tetrakis(triphenylphosphine)palladium(0) (Pd(PPh 3 ) 4 , 8 mg) is added. After another flushing with argon for 15 min, the reactant is heated to reflux for 32 h. Then the reactant is cooled down to room temperature, and extracted by Soxhlet extractor with methanol, hexane, and chloroform one by one. The polymer (93 mg, yield 96%) is recovered from the chloroform extract by precipitation in methanol and dried under vacuum. GPC: M n = 39.1 kDa; M w/ M n = 2.1. Anal. Calcd for C 28 H 36 F 2 N 2 OS (%): C, 69.10; H, 7.46; N, 5.76. Found (%): C, 68.08; H, 7.48; N, 5.71. 1 H NMR (CDCl 3 , 400 MHz): δ (p.p.m.) 8.81–7.72 (br, 3H), 4.89–4.03 (br, 2H), 2.43–0.53 (br, 31H).

General characterization 1 H NMR and 13 C NMR spectra of the corresponding compounds were measured on a Bruker DMX-400 spectrometer using d –chloroform as solvent and trimethylsilane as the internal reference. High-temperature gel permeation chromatography (GPC) measurements were carried out on Agilent PL-GPC 220 instrument, using 1,2,4-trichlorobenzene as the eluent at 160 °C. Thermogravimetric analysis (TGA) was measured on a Perkin-Elmer TGA-7 thermogravimetric analyzer with a heating rate of 10 °C min −1 under a nitrogen flow rate of 100 mL min −1 . UV–visible absorption spectra were measured on a Hitachi U-3010 UV–vis spectrophotometer. Electrochemical cyclic voltammograms was measured on a Zahner IM6e Electrochemical Workstation under a nitrogen atmosphere, with a Pt disk as working electrode, a Ag/AgCl as reference electrode, and a Pt wire as counter electrode in acetonitrile solution of tetrabutylammonium hexafluorophosphate ( n -Bu 4 NPF 6 ), and ferrocene/ferrocenium (Fc/Fc + ) couple was used as an internal reference.

Mobility measurements Hole mobilityterm·dom and electron mobility were measured using the space charge limited currentterm·mth (SCLCterm·mth) method. The hole-only device with the device structure of ITO /PEDOT: PSS/PTQ10term: IDICterm/Au was used to measure the hole mobilityterm·dom, and the electron-only device with the device structure of ITO/ZnO/PTQ10term: IDICterm/PDINO Al was used to measure the electron mobility. The hole and electron mobilities were calculated by MOTT–Gurney equation: where J is the current density, ε 0 the dielectric constant of empty space, ε r the relative dielectric constant of active layerterm·dom materials which is taken to be 3 in the calculation, μ the charge mobility, V the internal voltage in the device, and V = V appl − V bi − V s , where V appl is the voltage applied to the devices, V bi the built-in voltage from the relative work function difference between the two electrodes (0.2 V for the hole-only device and 0 V for the electron-only device), V s the voltage drop from the series resistance, and L the thickness of the active layersterm·dom.

GIWAXSterm·mth measurements GIWAXSterm·mth measurements were carried out using small angle X-ray scattering system (XEUSS, FRANCE Xenocs SAterm·mth). The samples for the GIWAXSterm·mth measurements were prepared on Si substrates using chloroform solutions of the samples. The 10 keV X-ray beam was incident at a grazing angle of 0.13–0.17°. The scattered X-rays were detected using a Dectris Pilatus 2M photon counting detector.

TEM characterization The TEM images were obtained on JEM-1011. The active layerterm·dom films for the TEM measurements were spin-coated onto ITO/PEDOT: PSS substrates, and the substrates with the active layersterm·dom were submerged in deionized water to make the active layersterm·dom fall off, then the active layerterm·dom films were picked up by copper grids for TEM measurements.

Device fabrication and characterization The PSCsterm·dom based on PTQ10term: IDICterm were fabricated with a device structure of ITO/PEDOT: PSS/PTQ10term: IDICterm/PDINO/Al. A thin layer of PEDOT: PSS was prepared on precleaned ITO glass through spin-coating a PEDOT: PSS aqueous solution (Baytron P VP AI 4083 from H. C. Starck) at 2000 rpm and dried subsequently at 150 °C for 15 min in air. Then the device was transferred to a glove box filled with nitrogen, in which the active layerterm·dom of PTQ10term: IDICterm was spin-coated from its chloroform solution onto the PEDOT: PSS layer at 3500 rpm. After spin-coating, the active layersterm·dom were annealed at 140 °C for 5 min for the devices with TAterm·mth treatment, and then the active layersterm·dom were treated by chloroform solvent for 30 s for the devices with solvent annealing treatment. The thickness of the active layerterm·dom is ca. 130 nm. Then methanol solution of PDINO at a concentration of 1.0 mg mL −1 was deposited upon the active layerterm·dom at 3000 rpm to afford a PDINO cathode buffer layer with thickness of ca. 10 nm. Finally, cathode metal Al was deposited onto the cathode buffer layer PDINO at a pressure of ca. 5.0 × 10 −5 Pa. The effective area of the devices was 4.7 mm 2 which was defined by Optical microscope (Olympus BX51). The current density–voltage ( J–V ) curves of the PSCsterm·dom were measured by scanning voltage from −1.5 V to 1.5 V with a voltage step of 10 mV and delay time of 1 ms on Keithley 2450 Source-Measure Unit in a glove box filled with nitrogen (oxygen and water contents are smaller than 0.1 ppm). Oriel Sol3A Class AAA Solar Simulator (model, Newport 94023A) with a 450 W xenon lamp and an air mass (AM) 1.5 filter was used as the light source. The light intensity was calibrated to be 100 mW cm −2 by a Newport Oriel 91150V reference cell. For accurately measuring the photocurrent, mask with an area of 2.2 mm 2 was used to define the effective area of the devices. The results with or without mask showed consistent values with relative errors within 0.5% (the devices with mask give slightly higher PCEterm·dom due to its slightly higher FFterm·dom). The PCEterm·dom results in the manuscript are obtained from the measurement without mask and PCEterm·dom statistics were obtained using more than 20 individual devices fabricated under the same conditions. The EQEterm·dom was measured by Solar Cell Spectral Response Measurement System QE-R3-011 (Enli Technology Co., Ltd., Taiwan). The light intensity at each wavelength was calibrated with a standard single-crystal Si photovoltaic cell.

Inverted devices were fabricated with a structure of ITO/ZnO/PTQ10term: IDICterm /MoO 3 /Ag. The ZnO precursor solution was prepared by dissolving 0.14 g of zinc acetate dehydrate (Zn(CH 3 COO) 2 ⋅2H 2 O, 99.9%, Aldrich) and 0.5 g of ethanolamine (NH 2 CH 2 CH 2 OH, 99.5%, Aldrich) in 5 ml of 2-methoxyethanol (CH 3 OCH 2 CH 2 OH, 99.8%, J&K Scientific). A thin layer of ZnO was deposited through spin-coating the ZnO precursor solution on precleaned ITO glass at 5000 rpm and baked subsequently at 200 °C for 30 min. Then the device was transferred into a glove box filled with nitrogen, in which the active layerterm·dom of PTQ10term: IDICterm (1: 1.5, w/w) was spin-coated from its chloroform solution onto the ZnO at 3500 rpm. After that, the active layersterm·dom were annealed at 140 °C for 5 min for the devices with TAterm·mth treatment, and treated by chloroform solvent for 30 s for the devices with solvent annealing treatment. The thickness of the active layerterm·dom is ca. 130 nm. Finally, a layer of ca. 5 nm MoO 3 and then a Ag layer of ca. 160 nm were evaporated subsequently under high vacuum.

Device stability measurements The inverted structured PSCsterm·dom with device structure of ITO/ZnO/PTQ10term: IDICterm /MoO 3 /Ag for stability measurements were encapsulated by ultraviolet-curable epoxy and thin glass slides and stored in nitrogen and air atmosphere, respectively.

Data availability The data that support the findings of this study are available from the corresponding author upon reasonable request.

Electronic supplementary material Supplementary Information Acknowledgements The work was supported by the Ministry of Science and Technology of China (973 project, No. 2014CB643501) and NSFC (Nos. 91633301, 91433117, 91333204, and 51673200) and the Strategic Priority Research Program of the Chinese Academy of Sciences, Grant No. XDB12030200. The authors would like to thank B.S. Runkun Huang from Institute of Chemistry, Chinese Academy of Sciences, for the help in TEM measurements, and to thank Dr. William Morrison from Molecular Vista, United States, for the help in the PiFMterm·mth measurements.

Author contributions C.S., Z.Z., and Y.L. designed the polymer PTQ10term, C.S. synthesized and characterized PTQ10term. F.P. carried out the PSCsterm·dom fabrication and characterization. H.B. and B.Q. participated in the discussion of the polymer synthesis. L.X. provided the cathode buffer layer material. J.Z. and Z.W. measured the GIWAXSterm·mth diffraction patterns and provided the device fabrication conditions for the inverted structured PSCsterm·dom. Y.L. supervised the project. C.S. and Y.L. wrote the paper. The first two authors contributed equally to this work.

Competing interests The authors declare no competing financial interests.

Footnotes Contributor Information Zhixiang WeipersName, Email: weizx@nanoctr.cn.

Zhi-Guo ZhangpersName, Email: zgzhangwhu@iccas.ac.cn.

YongfangpersName Li, Email: liyf@iccas.ac.cn.

References Associated Data Supplementary Materials Supplementary Information Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.