Side Chain Engineering on Medium Bandgap Copolymers to ...optics.nju.edu.cn/upload/20171222/201712221351484500.pdf · molecular weight sample by changing polymerization condi-tions

CommuniCation

1703344 (1 of 9) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advmat.de

Side Chain Engineering on Medium Bandgap Copolymers to Suppress Triplet Formation for High-Efficiency Polymer Solar Cells

Lingwei Xue, Yankang Yang, Jianqiu Xu, Chunfeng Zhang,* Haijun Bin, Zhi-Guo Zhang,* Beibei Qiu, Xiaojun Li, Chenkai Sun, Liang Gao, Jia Yao, Xiaofeng Chen, Yunxu Yang,* Min Xiao, and Yongfang Li*

DOI: 10.1002/adma.201703344

Over the past decade, polymer solar cells (PSCs) have attracted broad interest from tjhe academic and industrial communi-ties due to their unique advantages, such as light weight, low-cost production, the capability to be fabricated into flexible and semitransparent devices.[1–4] PSCs are composed of a blend active layer of p-type conjugated polymer as donor and n-type organic semiconductor (n-OS) as acceptor, sandwiched between a trans-parent bottom electrode and a metal top electrode.[5–7] Recently, power conversion efficiency (PCE) of the PSCs with medium bandgap conjugated polymer as donor and narrow bandgap n-OS as acceptor has reached 11–13%.[8–11] The key materials determining the photovoltaic performance of the PSCs are the conjugated polymer donors and the n-OS acceptors.

At present, the high performance acceptors are the narrow bandgap n-OS A–D–A type small molecules, such as ITIC (3,9-bis(2-methylene-(3-(1,1-dicyanomethylene)-indanone)-5,5,11,11-tetrakis(4-hexylphenyl)-dithieno[2,3-d:2,3′-d′]-s-indaceno[1,2-b:5,6-b′]-dithiophene)

Suppression of carrier recombination is critically important in realizing high-efficiency polymer solar cells. Herein, it is demonstrated difluoro-substitution of thiophene conjugated side chain on donor polymer can suppress triplet formation for reducing carrier recombination. A new medium bandgap 2D-conjugated D–A copolymer J91 is designed and synthesized with bi(alkyl-difluorothienyl)-benzodithiophene as donor unit and fluorobenzotriazole as acceptor unit, for taking the advantages of the synergistic fluorination on the backbone and thiophene side chain. J91 demonstrates enhanced absorption, low-lying highest occupied molecular orbital energy level, and higher hole mobility, in comparison with its control polymer J52 without fluorination on the thiophene side chains. The transient absorption spectra indicate that J91 can suppress the triplet formation in its blend film with n-type organic semiconductor acceptor m-ITIC (3,9-bis(2-methylene-(3-(1,1-dicyanomethylene)-indanone)-5,5,11,11-tetrakis(3-hexylphenyl)-dithieno[2,3-d:2,3′-d′]-s-indaceno[1,2-b:5,6-b′]-dithiophene). With these favorable properties, a higher power conversion efficiency of 11.63% with high VOC of 0.984 V and high JSC of 18.03 mA cm−2 is obtained for the polymer solar cells based on J91/m-ITIC with thermal annealing. The improved photovoltaic performance by thermal annealing is explained from the morphology change upon thermal annealing as revealed by photoinduced force microscopy. The results indicate that side chain engineering can provide a new solution to suppress carrier recombination toward high efficiency, thus deserves further attention.

Polymer Solar Cells

L. Xue, Y. Yang, H. Bin, Prof. Z.-G. Zhang, B. Qiu, X. Li, C. Sun, L. Gao, J. Yao, X. Chen, Prof. Y. F. LiCAS Research/Education Center for Excellence in Molecular Sciences CAS Key Laboratory of Organic Solids, Institute of ChemistryChinese Academy of SciencesBeijing 100190, ChinaE-mail: [email protected]; [email protected]. XueCollege of Chemistry and Chemical EngineeringPingdingshan UniversityPingdingshan, Henan 467000, ChinaJ. Xu, Prof. C. F. Zhang, M. XiaoNational Laboratory of Solid State Microstructures, School of Physics and Collaborative Innovation Center of Advanced MicrostructuresNanjing UniversityNanjing 210093, China E-mail: [email protected]

Y. Yang, H. Bin, X. Li, C. Sun, L. Gao, Prof. Y. F. LiSchool of Chemistry and Chemical EngineeringUniversity of Chinese Academy of SciencesBeijing 100049, ChinaProf. Y. X. YangDepartment of Chemistry & Chemical EngineeringSchool of Chemical & Biological EngineeringUniversity of Science & Technology BeijingBeijing 100083, ChinaE-mail: [email protected]. Xu, Prof. C. F. Zhang, M. XiaoSynergetic Innovation Center in Quantum Information and Quantum PhysicsUniversity of Science and Technology of ChinaHefei, Anhui 230026, China

Adv. Mater. 2017, 29, 1703344

© 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1703344 (2 of 9)

www.advmat.dewww.advancedsciencenews.com

and IDIC (2,7-bis(2-methylene-(3-(1,1-dicyanomethylene)-indanone)-4,4,9,9-tetrahexyl-4,9-dihydro-s-indaceno[1,2-b:5,6-b']dithiophene).[12,13] Especially, m-ITIC[14] by the side chain isomerization of ITIC (as shown in Figure 1b) possesses a stronger film absorption coefficient and higher electron motility than that of ITIC, demonstrating the improved pho-tovoltaic performance of the nonfullerene PSCs.[15,16] Along with the development of the n-OS acceptors, medium bandgap polymer donors are specially required to afford complementary absorption for efficiently harvesting solar photons.[17] Recently, we developed a series of medium bandgap D–A copolymers based on bithienylbenzodithiophene (BDTT)-alt-benzotriazole (BTA) backbone with different conjugated side chains.[8,18–20] With fluorination on the BTA units and with the side chain engineering[21] (alkylthio or trialkylsilyl group), the resultant polymers demonstrated high PCEs of 9−11% in the non-fullerene PSCs with low bandgap n-OS ITIC as acceptor.[8,18–20] The introduction of fluorine substituent has been proven to be an effective strategy of producing ordered crystalline struc-ture for a higher hole mobility through intra- and/or inter-molecular F⋯H and F⋯S interactions.[22,23] Furthermore, the electronegative effect of the fluorine atom can cause a down-shift of the HOMO (highest occupied molecular orbital) level for a higher open-circuit voltage (VOC) of the PSCs.[24–26] As a typical example, the polymer J51 (Scheme 1) obtained by the fluorination of J50 possesses an improved hole mobility and low-lying HOMO energy level of −5.26 eV (EHOMO = −5.1 eV for J50). The improved photophysical properties result in signifi-cantly improved photovoltaic performance, yielding a high PCE

Prof. Y. F. LiLaboratory of Advanced Optoelectronic Materials College of Chemistry Chemical Engineering and Materials ScienceSoochow UniversitySuzhou, Jiangsu 215123, China

of 9.26% for J51 along with an increased VOC of 0.82 V (PCE = 4.80%, VOC = 0.71 V for J50).[19]

With the success of the fluorination strategy, herein, we further introduced two fluorine substituents on the thiophene conjugated side chains and synthesized a new polymer J91 with the conjugated and fluorinated alkyl(3,4-difluoro)thienyl side chains. Compared with its polymer analog J52, J91 with the difluoro-substituents on the thiophene conjugated side chains showed further down-shifted HOMO energy level at −5.50 eV and an enhanced ordered crystalline structure. The optical, electrochemical, and photovoltaic properties of J91 were investi-gated, and compared with those of its counterpart polymer J52. Interestingly, with the alkyldifluorothienyl side chain, J91 dem-onstrated suppression of triplet formation for reducing carrier recombination in its blend film with m-ITIC, as disclosed by the transient absorption spectroscopic measurements. There-fore, a higher PCE of 11.63% with high VOC of 0.984 V and high JSC of 18.03 mA cm−2 was obtained for the PSCs based on J91/m-ITIC.

The synthetic routes and chemical structure of J91 are depicted in Scheme 1. The conversion of 3,4-dibromothiophene (1) to compound 3 was accomplished initially by lithiation with lithium diisopropylamide (LDA) and treatment with 2-ethyl-hexyl bromide, followed by lithiation with LDA again and by the treatment with trimethylchlorosilane. Alkyl(3,4-difluoro)thio-phene (4) was obtained by lithiation of compound 3 with n-BuLi (n-Butyllithium) followed by the treatment with (PhSO2)2NF (N-fluorobenzenesulfonimide) and TBAF (tetrabutyl-ammonium fluoride).[27] The distannyl BDTFT (3,4-difluoro-alkylthienyl benzo[1,2-b:4,5-b']dithiophene) monomer was readily prepared with high yield via a nucleophilic attack on 5 with a lithiated compound 4 precursor, followed by aromatiza-tion with hydrochloric acid SnCl2 solution and subsequently lithiation with n-BuLi followed by the treatment with SnMe3Cl. As shown in Scheme 1, the copolymer J91 was synthesized

Adv. Mater. 2017, 29, 1703344

Figure 1. Side chain engineering on polymer donors and n-OS acceptors toward high efficiency.



by BDTFT unit with the FBTA (4,7-bis(5-bromothiophen-2-yl)-5,6-difluoro-2-(2-hexyldecyl)-2H-benzo[d][1,2,3]triazole) unit using the palladium-catalyzed Stille cross-coupling reaction. Also, J52 was synthesized following the procedure described previously,[20] for a comparative study of the effect of difluoro-substitution of thiophene conjugated side chains on their photo physical properties.

The obtained copolymer was precipitated in methanol and then purified by Soxhlet extraction to remove the oligomers and other impurities. The copolymers showed good solubility in chloroform, chlorobenzene, and o-dichlorobenzene at room temperature. The weight average molecular weight (Mw), number-average molecular weight (Mn), and polydispersity index (PDI, Mw/Mn) of the polymers were measured using gel permeation chromatography against polystyrene standards in o-dichlorobenzene at 160 °C with the results listed in Table 1. The molecular weight Mn of J91 is 6.97 kDa which is relatively low in comparison with that of J52. We tried to get higher molecular weight sample by changing polymerization condi-tions including prolonging reaction time from 12 to 48 h, using mixed solvent of toluene/DMF (N,N-dimethylformamide). However, changing the polymerization conditions cannot help to get higher molecular weight for J91. The low Mn value of J91 could be due to the low reaction activity between the alkyl(3,4-difluoro)thiophene substituted BDTFT unit and the difluoro-substituted FBTA unit. Probably, the photovoltaic performance of J91 could be further improved if high molecular weight pol-ymer can be obtained.

To evaluate the thermal property of J91, thermogravimetric analysis was performed, as shown in Figure S1 in Supporting

Information. The copolymer J91 displayed good thermal sta-bility with the 5% weight-loss temperature at 385 °C under nitrogen flow, which is much higher than that (312 °C) of J52. The result indicates that thermal stability of the polymer J91 with fluorination on the side chains is improved.

Figure 2a shows the absorption spectra of solid films of J52 and J91. It can be seen that the two polymer films pos-sess similar absorption profile. But, the absorption coeffi-cient (0.98 × 105 cm−1) of J91 film at 536 nm is significantly enhanced than that (0.74 × 105 cm−1) of J52 film, due probably to the existence of more ordered aggregation and stronger π–π stacking interaction in the J91 polymer film. Notably, the high film absorption coefficient of J91 is comparable to that of m-ITIC acceptor (1.04 × 105 cm−1, as also shown in Figure 2a for comparison), and the complementary absorp-tion of the polymer donor and m-ITIC acceptor will be bene-ficial to enhance light-harvest and to increase short-circuit current density of the PSCs.

To understand the effect of the fluorinated side chains on the physicochemical properties of J91, its electronic energy level was measured by electrochemical cyclic voltammetry and compared with that of J52. From the onset oxidation (ϕox) and onset reduction (ϕred) potentials of the cyclic voltammo-grams, as shown in Figure 2b, the HOMO and LUMO (lowest unoccupied molecular orbital) energy levels were calculated according to the equations: EHOMO = −e(ϕox + 4.36) (eV) and ELUMO = −e(ϕred + 4.36) (eV), where the unit of potential is V versus Ag/AgCl.[8] The ϕox/ϕred values are 1.14/−1.34 V for J91, and 0.81/−1.37 V for J52. Then, their HOMO/LUMO energy levels are −5.50/−3.02 eV for J91 and −5.17/−2.99 eV for J52. The 0.33 eV lower-lying EHOMO of J91 than that of J52 could be ascribed to the strong electronegative effect of the bifluoro-substituents in its thiophene conjugated side chains, which will benefit for higher VOC of the PSCs with J91 as donor. Recently, Fan et al. introduced mono-fluorine substituent on the thio-phene conjugated side chains of J52, showing a lower HOMO energy level of −5.36 eV.[28] Our result shows that difluorine substitution on the conjugated side chains is more effective in lowering the HOMO energy level of the polymers.

Adv. Mater. 2017, 29, 1703344

Scheme 1. Synthesis routes of the copolymer J91.

Table 1. Molecular weights and physicochemical properties of J52 and J91.

Polymers Mn [kDa]

Mw [kDa]

Mw/Mn Egopt

[eV]HOMO

[eV]LUMO [eV] Eg

ec [eV]

Td [°C]

J52 57.5 106.95 1.86 1.96 −5.17 −2.99 2.18 312

J91 6.97 13.40 1.92 2.00 −5.50 −3.02 2.48 385



Photovoltaic performance of the polymers was studied by fabricating the PSCs based on the polymers as donor and m-ITIC as acceptor, with a conventional device structure of ITO (indium tin oxide)/PEDOT:PSS (poly(3,4-ethylenedioxy-thiophene): poly(styrene-sulfonate))/Polymer donor: m-ITIC/PDINO (perylene diimide functionalized with amino N-oxide)/Al. In the device fabrication, the polymer donor: m-ITIC weight ratio of 1:1.5 was adopted according to our previous studies.[14] Chloroform was used as the processing solvent because it afforded the best-quality blend films. And thermal annealing of the active layers at 150 °C for 2 min was performed to optimize the photovoltaic performance of the PSCs.

Figure 3a shows the current density–voltage (J–V) curves of the PSCs as cast and with the thermal annealing at 150 °C for 2 min under the illumination of AM 1.5 G, 100 mW cm−2, and

Table 2 lists the VOC, JSC, fill factor (FF), and PCEs of the PSCs. It can be seen that thermal annealing improved the photo-voltaic performance significantly. For the PSCs without thermal annealing (as cast), the PCE values are 3.45% and 6.05% for the devices with J52 and J91 as donor, respectively. With the thermal annealing treatment at 150 °C for 2 min, the PCE of the PSC based on J91: m-ITIC reached a high value of 11.63% with a high Voc of 0.984 V, JSC of 18.03 mA cm−2, and FF of 65.54%, while the PCE of the PSC based on J52: m-ITIC was also improved to 5.98% with a Voc of 0.701 V, JSC of 17.16 mA cm−2, and FF of 49.73%. The high Voc of 0.984 V for the PSCs based on J91:m-ITIC should be ascribed to the lower-lying HOMO level of the polymer donor J91 with the bifluoro-thienyl side chains. Notably, the PCE of 11.63% is one of the highest values in the PSCs. The excellent photovoltaic performance of J91

Adv. Mater. 2017, 29, 1703344

Figure 2. a) Film absorption coefficient of J52, J91, and m-ITIC in film state. b) Cyclic voltammograms of J52 (blue line) and J91 (red line) films on glassy carbon electrode in 0.1 mol L−1 Bu4NPF6 acetonitrile solution at a potential scan rate of 100 mV s−1; the inset shows the cyclic voltammogram of ferrocene/ferrocenium (Fc/Fc+) couple used as an internal reference.

Figure 3. a) J–V characteristics of the PSCs based on polymer donor/m-ITIC (1:1.5, w/w) with or without thermal annealing (TA) at 150 °C for 2 min, under the illumination of AM 1.5 G, 100 mW cm−2. b) The corresponding IPCE spectra of the PSCs. c) VOC and Eloss versus EHOMO. d) Light intensity dependence of JSC of the devices.



indicates that the bifluoro-substitution on the thiophene conju-gated side chains is an effective way to further improve photo-voltaic performance of the conjugated polymers with thiophene conjugated side chains.

Figure 3b shows the input photon to converted current effi-ciency (IPCE) spectra of the PSCs. The IPCE curves showed a broad photoresponse from 300–800 nm, benefitted from the complementary absorption of the donors (J52 and J91) and m-ITIC acceptor. And the IPCE values were increased for the PSCs with thermal annealing treatment. The calculated Jsc by integrating the IPCE curve with an AM 1.5 G reference spec-trum is 17.24 mA cm−2 for the PSC based on J91:m-ITIC with thermal annealing, which is within 5% mismatch in comparison to the corresponding JSC value obtained from the J–V curves. The result indicates that our measurement is highly reliable.

From Figure 3a and Table 2, it can be seen that one dramatic effect of the difluorine substitution of thiophene conjugated side chains on the photovoltaic properties is the increase of VOC, which is benefitted from the down-shifted HOMO energy level of the polymer donor. A high Voc value of ≈0.98 V was achieved for the PSC with J91 as donor with the difluorothienyl conju-gated side chains relative to that (0.70 V) of the device based on control polymer J52 without the two fluorine substitution on the side chains.

Figure 3c shows the relationship between the VOC and the EHOMO of our reported fluorobenzotriazole-based medium bandgap polymers with different conjugated side chains including J50,[19] J51,[19] J61,[14,20] J71,[8] J8116 together with the polymers of J91 and J52 reported in this work. It can be seen that VOC increased with the decrease of the EHOMO values of the donor, and J91 delivered the highest VOC value. Figure 3c also displays the dependence of photon energy loss (Eloss) of the PSCs on the EHOMO of the polymer donors. Eloss is defined as Eloss = Eg − eVoc, where Eg is the lowest optical bandgap of the donor and acceptor components.[29–31] It can be seen that the Eloss values of the PSCs with the J-series polymers as donor decreased with the down-shifting of their HOMO energy level, and the J91-based device shows the lowest Eloss of 0.61 eV. The results indicate that the difluoro-substitution on the thiophene conjugated side chains is an effective way to further down-shift the EHOMO of the polymer donor for achieving a higher VOC value and decreasing Eloss of the devices. Notably, the HOMO energy difference (ΔEHOMO) between J91 donor and m-ITIC acceptor (EHOMO = −5.54 eV).[14] is only 0.04 eV, which is a big challenge for achieving high-efficiency PSCs due to the excitonic nature of organic semiconductors. Interestingly, the device still demonstrated a high JSC of 18.03 mA cm−2 along with a high VOC of 0.98 V.

To understand the thermal annealing effect on the device performance, the bulk charge transport proper-ties of the polymer/m-ITIC blends were investigated using the space charge limited current method with the hole only device (ITO/PEDOT:PSS/active layer/Au) and elec-tron only device (ITO/ZnO/active layer/PDINO/Al). The plots of the current density versus voltage for the devices are shown in Figure S2 in the Supporting Information. The hole (µh)/electron (µe) mobilities of the J52/m-ITIC film are calculated to be 7.415 × 10−6 cm2 V−1 s−1/2.284 × 10−4 for the as-cast film and 2.094 × 10−5 cm2 V−1 s−1/2.376 × 10−4 cm2 V−1 s−1 for the thermal-annealed film. And the µh/µe mobilities of the J91/m-ITIC film are calculated to be 1.161 × 10−5 cm2 V−1 s−1/2.583 × 10−4 cm2 V−1 s−1 for the as-cast film and 1.016 × 10−4 cm2 V−1 s−1/3.002 × 10−4 cm2 V−1 s−1 for the thermal-annealed film. Apparently, after thermal annealing, charge carrier mobilities were increased and more balanced to some extent. The increased and more balanced charge car-rier mobilities correlate well with the increased FF values of the PSCs after thermal annealing. In addition, in comparison with the charge carrier mobilities of the J52-based devices, the devices based on J91 demonstrated higher hole mobilities, which could be due to the enhanced intramolecular or intermo-lecular F⋯H and F⋯S interactions by the difluoro-substitution on the thiophene conjugated side chains.

To further understand the effect of fluorinated side chains and thermal annealing on the photovoltaic properties, light intensity dependent photocurrent measurements were con-ducted to compare charge recombination behavior of the devices. In general, JSC has a power−law dependence on the irradiated light intensity (Plight): JSC ∝ Plight

α, where α should be unity when the bimolecular recombination is negligible. As shown in Figure 3d, for the as-cast devices, the fitted α values are 0.897 for the J52: m-ITIC-based device and 0.925 for the J91: m-ITIC-based device, while remarkably higher α values of 0.959 and 1.028 were observed for the corresponding annealed devices, respectively. The results indicate that both device engi-neering (with thermal annealing) and side chain engineering (with the fluorinated side chains) can suppress the charge recombination in the devices. And the suppressed charge recombination correlates well with the more balanced carrier mobilities mentioned above and should partially account for the higher fill factors for the annealed devices.

In the PSCs, the active layer absorbs photons to create exci-tons, then the excitons diffuse to the donor/acceptor interfaces where they dissociate into a charge transfer (CT) state electron–hole pair with electron in the LUMO of the acceptor and hole in the HOMO of the donor at the interface. The electron–hole

Adv. Mater. 2017, 29, 1703344

Table 2. Photovoltaic performance parameters of the PSCs based on polymer: m-ITIC (1:1.5, w/w), under the illumination of AM 1.5G, 100 mW cm−2.

Devicesa) Voc [V]

Jsc [mA cm−2]

FF [%]

PCE [%]

Rsh [KΩcm−2]

Rs [Ω cm−2]

Jscc)

[mA cm−2]

J52:m-ITIa) 0.701 12.03 40.96 3.45 0.13 16.77 –

J91:m-ITIa) 0.979 14.43 42.86 6.05 0.23 20.02 –

J52:m-ITIb) 0.701 (0.699 ± 0.003) 17.16 (17.24 ± 0.29) 49.73 (48.90 ± 0.61) 5.98 (5.89 ± 0.12) 0.29 (0.27 ± 0.02) 10.21 (11.12 ± 0.63) 16.33

J91:m-ITIb) 0.984 (0.984 ± 0.003) 18.03 (17.73 ± 0.32) 65.54 (64.90 ± 0.78) 11.63 (11.32 ± 0.21) 1.44 (1.19 ± 0.20) 9.18 (10.15 ± 0.49) 17.24

a)As-cast film; b)Thermal annealing at 150 °C for 2 min and the values in parentheses are average values obtained from 15 devices; c)Calculated from IPCE.



pairs at CT state could be further dissociated into the charge separated (CS) free carriers that can contribute to the photocur-rent. The processes of germinate and biomolecular recombina-tion compete against the charge transfer and charge separation processes, setting the efficiency limit of the PSCs.[32,33]

To understand the mechanism underlying the devices studied in this work, we performed fs-resolved and ns-resolved transient absorption (TA) spectroscopy measurements with probe light covering from visible to near infrared range (see the Experimental Section in the Supporting Information). Figure 4a displays typical TA spectra recorded from the neat films of donor and acceptor, showing the main features of ground-state bleaching (GSB) and excited-state absorption (ESA). GSB signals in both films of the donor J91 and the acceptor m-ITIC appear in the spectral ranges close to their major absorption bands (Figure 2a). Polaron-induced ESA signal appears with a broad band feature in the infrared range of 800–1400 nm in the J91 film and with a sharp peak centered at 960 nm in the m-ITIC film, respectively.

Figure 4b shows the TA spectra in the annealed film of J91/m-ITIC blend recorded at different time delays. As marked in Figure 4b, GSB and ESA signals of donor and acceptor com-pounds are observed in the spectral range close to those in the neat films. Following the decay of these signals, a photo-induced absorption signal gradually builds up in the infrared range centered at 1220 nm, as a signature of the formation of excited states in the blends that are not present in neat films. We compare the kinetics in the blend film with those in neat films to elucidate the interfacial charge dynamics. At the early stage (Figure 4c), the decay lifetime of GSB signal probed at

595 nm in the blend film is dramatically shortened to ≈0.2 ps in comparison with that in the J91 film, which can be ascribed to the electron transfer from J91 to m-ITIC. In the time scale of 10–100 ps, the GSB signal shows an unexpected growth which is possibly caused by the hole transfer to J91 from m-ITIC excited by 500 nm pump. The assignment of hole transfer is further confirmed by the hole dynamics in the blend film when the pump wavelength is set at 710 nm to selectively excite m-ITIC. The decay of GSB signal probed at 720 nm in the blend film becomes significantly faster than that in the m-ITIC film (Figure 5d). The early-stage decay curve consists of two exponential components with characteristic lifetime parameters of ≈0.13 ps and ≈3.4 ps (Figure 4d), while the GSB signal at 595 nm is simultaneously built up within a similar temporal scale (Figure 4d). These results strongly support a high effi-ciency of hole transfer from photoexcited m-ITIC to J91.[33] The coexistence of electron and hole transfer in J91/m-ITIC blends confirms the promising potential of non-fullerene PSC over conventional fullerene-based PSC with the advantage of con-verting the light absorbed by the acceptor.

The CT state excitons compete against germinate recombi-nation to dissociate into free charges. The GSB bleach signals in the blend film persist to the time scale much longer than exciton lifetimes in neat films (Figure 4c,d), suggesting a high efficiency of charge separation where the bimolecular recombi-nation is a major competing process. Simultaneously following the recovery of the ESA signals of photoexcited polarons in the time scale of 100–1000 ps, the gradual buildup of ESA band centered at 1220 nm suggests the formation of a new species during the bimolecular recombination. It is more important

Adv. Mater. 2017, 29, 1703344

Figure 4. Charge dynamics in the J91/m-ITIC blend film. a) Transient absorption (TA) spectra recorded from neat films of J91 and m-ITIC, respectively. b) TA spectra recorded from the J91/m-ITIC blend film at different time delays. c) The kinetic curves of the J91/m-ITIC blend film probed at 595 nm in comparison with that of the neat J91 film with pump at 500 nm. d) The kinetic curves of the J91/m-ITIC blend film probed at 720 and 595 nm in comparison with that recorded at 720 nm for the neat J91 film with pump at 710 nm. e) The fs-resolved TA curves probed at 1220 nm comparing the dynamics of triplet formation in annealed and unannealed blend films. f) The ns-resolved TA curves probed at 1220 nm comparing the decay dynamics of triplet formation in annealed and unannealed blend films.



that the ESA signal at 1220 nm is highly correlated with device performance: The signal amplitude is weaker in the blend with higher device efficiency (Figure 4e). In the thermal-annealed sample of the J91/m-ITIC film, the amplitude of late-stage signal at 1220 nm is much less than that in the unannealed samples despite that the polaron-induced ESA signal is larger in the annealed sample (Figure 4e,f). In analogy to carrier dynamics in fullerene-based OPV blends,[34,35] the ESA signal at 1220 nm is probably caused by the triplet formation during the bimolecular recombination of charge separated excitons. We further perform the control experiment with triplet sensitizer (Figure S3 in the Supporting Information). The results suggest that the ESA signal at 1220 nm is caused by the triplet forma-tion in m-ITIC, which strongly support our explanation.

With the above results, we can tentatively summarize the carrier dynamics responsible for the device performance (Figure 5). At the interfaces, the early-stage charge dynamics forms the CT state excitons by the processes of either electron transfer from J91 to m-ITIC or hole transfer from m-ITIC to J91 (Figure 5a). The dissociation of CT state excitons into CS free carriers contributes to the photocurrent generation (Figure 5b), which, however, is competed against the process of bimolecular recombination with the formation of triplets. The process of triplet formation is strongly suppressed and charge recombina-tion is significantly reduced after thermal-annealing, resulting in a higher current density in the devices. The results suggest a critical role played by the nanoscale morphology to the carrier

dynamics (Figure 4), which is also supported by a compara-tive study on the blend films of J91/m-ITIC and J52/m-ITIC (Figure 5c,d). The triplet formation is much stronger in J52/m-ITIC with a much longer lifetime of GSB signal than that in J91/m-ITIC, which is consistent with the lower efficiency in the J52/m-ITIC devices. In spite of a larger triplet signal, the onset of ESA at 1220 nm starts at a later time in the J52/m-ITIC film, suggesting the process of charge separation is possibly slower. These results are reasonable since the fluorine atoms in the side chains may strongly modify the film morphology that is responsible for the difference in carrier dynamics.[31] In other words, the superior photovoltaic performance of J91/m-ITIC devices over J52/m-ITIC devices can be explained by the rela-tively higher open-circuit voltage and the suppression of triplet formation and charge recombinaiton in J91/m-ITIC.

As is well known, photovoltaic performance of the PSCs depends sensitively on their nanoscale morphology. Here the thermal annealing effect on the morphology of the active layer was examined with an emergent technology, photoinduced fore microscopy (PiFM).[36] By imaging at characteristic Fou-rier transform infrared (FT-IR) wavelengths corresponding to absorption peaks of donor and acceptor chemical species, PiFM has demonstrated the ability to spatially map nm-scale patterns of the individual chemical components in their blend films.[37,38] Thus, this technology can well address the correla-tion between the spatial chemical information with the mor-phology of nanostructures.

Adv. Mater. 2017, 29, 1703344

Figure 5. State diagrams representing the a) charge transfer and b) charge separation dynamics in the donor/acceptor blends of PSCs. c) The dynamics of triplet formation in the blend films of J91/m-ITIC and J52/m-ITIC, respectively. d) The ns-TA kinetic curves probed at 595 nm for the blend films of J91/m-ITIC and J52/m-ITIC, respectively.



Adv. Mater. 2017, 29, 1703344

Figure 6 provides the topography PiFM images of the as-cast and thermal annealed blend films along with their cor-responding phase images. Obviously, the conventional atomic force microscopy images have a limitation in discriminating the spatial distribution of the donor and acceptor components and shedding light on the effect of thermal annealing. As shown in Figure 6e, J91 and m-ITIC have characteristic FT-IR absorption peaks, associated with their chemical structure. By imaging at 1701 cm−1 for m-ITIC and 816 cm−1 for J91 with PiFM, spatially mapped patterns of m-ITIC and J91 in their blend films are clearly visible. Notably, the wavenumber used for imaging J91 (816 cm−1) is slightly off the FTIR absorption peak (801 cm−1) but is still effective in discriminating them against the m-ITIC molecules since there is much less absopr-tion for m-ITIC at that wavenumber. As depicted in Figure 6f, the as-cast J91: m-ITIC films have good miscibility between the two component. Although the well-matched distribution of m-ITIC molecules over J91 networks can offer large hetero-junction interface area for exciton dissociation, proper phase separation is desirable in the composite films to enable higher charge carrier mobilities and a continuous network for hole and electron transport to the corresponding electrodes. When thermal annealing was applied (Figure 6g), proper phase sepa-ration with an average domain size of ≈25 nm was observed. It has been reported that the intra- and/or intermolecular F⋯H and F⋯S interactions in organic semiconductors can also increase its crystallization.[19] As for J91, its high fluorination on the backbone and side chains increases its crystallization. This crystallization should strongly influence bulk-heterojunc-tion phase separation upon thermal annealing. Notably, the observed domain size is very close to the typical exciton diffu-sion length. This ideal phase separation behavior facilitates an improved exciton dissociation and carrier collection efficiency,

thus leading to an increase in FF and Jsc and affording higher device efficiency. This work also suggests the significant poten-tial of PiFM to explore the phase separation behavior of the composite films at the nanoscale.

In conclusion, to take the advantages of the synergistic fluorination effect of the backbone of the BTA units and side chain on BDT units, a new highly fluorinated BDT-alt-BTA polymer J91 was synthesized with introducing alkyl(3,4-dif-luoro)thienyl side groups into the BDT donor units. Compared with the control polymer J52 without fluorine substituent on its side chains, J91 with alkyl-difluorothienyl side chains exhibits more intense absorption, low-lying HOMO energy level, and higher charge carrier mobility. More interestingly, J91 demonstrated suppression of triplet formation in its blend film with m-ITIC as disclosed by the transient absorption (TA) spectroscopic measurements, which indicates the reduced charge recombination in the J91/m-ITIC blend films. These favorable properties translate into significantly high efficiency of 11.63% with simultaneously high VOC of 0.984 V and high JSC of 18.03 mA cm−2. Notably, the PCE of 11.63% is one of the highest values achieved in PSCs. In addition, the effect of thermal-annealing on the active layer morphology and photo-voltaic performance was investigated by transient absorption measurement and PiFM analysis. The results indicate that introducing alkyldifluorothienyl side chains to suppress the triplet formation and charge recombination is an effective way for further improving photovoltaic performance of the high-efficiency conjugated polymer donors for PSCs.

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

Figure 6. Topography images of a) the as-cast and c) thermal annealed J91: m-ITIC blend films along with b,d) their corresponding phase images. e) FT-IR spectra of J91 (green line) and m-ITIC (red line). PiFM images of f) the as-cast and g) thermal annealed J91: m-ITIC blend films examined at 816 and 1701 cm−1 corresponding to the absorption bands for J91 and m-ITIC, respectively. The spectral response for J91 component is marked with green and with red for m-ITIC.



Adv. Mater. 2017, 29, 1703344

AcknowledgementsThe work was supported by the Ministry of Science and Technology of China (973 project, Grant No. 2014CB643501) and NSFC (Grant Nos. 91633301, 91433117, 91333204, 11574140, 21374124, 51722308 and 51673200) and the Strategic Priority Research Program of the Chinese Academy of Sciences, Grant No. XDB12030200.

Conflict of InterestThe authors declare no conflict of interest.

Keywordsfluorination effect, polymer donors, photoinduced force microscopy, polymer solar cells

Received: June 15, 2017Revised: July 23, 2017

Published online: August 31, 2017

[1] G. Yu, J. Gao, J. C. Hummelen, F. Wudl, A. J. Heeger, Science 1995, 270, 1789.

[2] J. Chen, Y. Cao, Acc. Chem. Res. 2009, 42, 1709.[3] Y.-J. Cheng, S.-H. Yang, C.-S. Hsu, Chem. Rev. 2009, 109, 5868.[4] G. Li, R. Zhu, Y. Yang, Nat. Photonics 2012, 6, 153.[5] B. C. Thompson, J. M. J. Fréchet, Angew. Chem., Int. Ed. 2008, 47,

58.[6] Y. F. Li, Acc. Chem. Res. 2012, 45, 723.[7] H. Yao, L. Ye, H. Zhang, S. Li, S. Zhang, J. Hou, Chem. Rev. 2016,

116, 7397.[8] H. Bin, L. Gao, Z.-G. Zhang, Y. Yang, Y. Zhang, C. Zhang, S. Chen,

L. Xue, C. Yang, M. Xiao, Y. F. Li, Nat. Commun. 2016, 7, 13651.[9] S. Li, L. Ye, W. Zhao, S. Zhang, S. Mukherjee, H. Ade, J. Hou, Adv.

Mater. 2016, 28, 9423.[10] F. Zhao, S. Dai, Y. Wu, Q. Zhang, J. Wang, L. Jiang, Q. Ling, Z. Wei,

W. Ma, W. You, C. Wang, X. Zhan, Adv. Mater. 2017, 22, 1700144.[11] W. Zhao, S. Li, H. Yao, S. Zhang, Y. Zhang, B. Yang, J. Hou, J. Am.

Chem. Soc. 2017, 139, 7148.[12] Y. Lin, J. Wang, Z.-G. Zhang, H. Bai, Y. F. Li, D. Zhu, X. Zhan, Adv.

Mater. 2015, 27, 1170.[13] Y. Lin, Q. He, F. Zhao, L. Huo, J. Mai, X. Lu, C.-J. Su, T. Li, J. Wang,

J. Zhu, Y. Sun, C. Wang, X. Zhan, J. Am. Chem. Soc. 2016, 138, 2973.[14] Y. Yang, Z.-G. Zhang, H. Bin, S. Chen, L. Gao, L. Xue, C. Yang,

Y. F. Li, J. Am. Chem. Soc. 2016, 138, 15011.

[15] S. Chen, H. J. Cho, J. Lee, Y. Yang, Z.-G. Zhang, Y. F. Li, C. Yang, Adv. Energy Mater. 2017; https://doi.org/10.1002/aenm.201701125.

[16] H. Bin, L. Zhong, Y. Yang, L. Gao, H. Huang, C. Sun, X. Li, L. Xue, Z.-G. Zhang, Z. Zhang, Y. F. Li, Adv. Energy Mater. 2017, http://dx.doi.org/10.1002/aenm.201700746.

[17] Y. Cai, L. Huo, Y. Sun, Adv. Mater. 2017, 29, 1605437.[18] J. Min, Z.-G. Zhang, Y. Hou, C. O. Ramirez Quiroz, T. Przybilla,

C. Bronnbauer, F. Guo, K. Forberich, H. Azimi, T. Ameri, E. Spiecker, Y. F. Li, C. J. Brabec, Chem. Mater. 2015, 27, 227.

[19] L. Gao, Z.-G. Zhang, H. Bin, L. Xue, Y. Yang, C. Wang, F. Liu, T. P. Russell, Y. F. Li, Adv. Mater. 2016, 28, 8288.

[20] H. Bin, Z.-G. Zhang, L. Gao, S. Chen, L. Zhong, L. Xue, C. Yang, Y. F. Li, J. Am. Chem. Soc. 2016, 138, 4657.

[21] Z.-G. Zhang, Y. F. Li, Sci. China: Chem. 2015, 58, 192.[22] S. C. Price, A. C. Stuart, L. Yang, H. Zhou, W. You, J. Am. Chem. Soc.

2011, 133, 4625.[23] W. Li, S. Albrecht, L. Yang, S. Roland, J. R. Tumbleston, T. McAfee,

L. Yan, M. A. Kelly, H. Ade, D. Neher, W. You, J. Am. Chem. Soc. 2014, 136, 15566.

[24] M. Zhang, X. Guo, S. Zhang, J. Hou, Adv. Mater. 2014, 26, 1118.[25] D. Liu, W. Zhao, S. Zhang, L. Ye, Z. Zheng, Y. Cui, Y. Chen, J. Hou,

Macromolecules 2015, 48, 5172.[26] J. Wolf, F. Cruciani, A. El Labban, P. M. Beaujuge, Chem. Mater.

2015, 27, 4184.[27] S.-N. Cho, W.-H. Lee, J. B. Park, J.-H. Kim, D.-H. Hwang, I.-N. Kang,

Synth. Met. 2015, 210, 273.[28] Q. Fan, W. Su, X. Meng, X. Guo, G. Li, W. Ma, M. Zhang, Y. F. Li,

Solar RRL 2017, 1, 1700020.[29] D. Veldman, S. C. J. Meskers, R. A. J. Janssen, Adv. Funct. Mater.

2009, 19, 1939.[30] M. A. Faist, T. Kirchartz, W. Gong, R. S. Ashraf, I. McCulloch,

J. C. de Mello, N. J. Ekins-Daukes, D. D. C. Bradley, J. Nelson, J. Am. Chem. Soc. 2012, 134, 685.

[31] K. Kawashima, Y. Tamai, H. Ohkita, I. Osaka, K. Takimiya, Nat. Commun. 2015, 6, 10085.

[32] T. M. Clarke, J. R. Durrant, Chem. Rev. 2010, 110, 6736.[33] O. Ostroverkhova, Chem. Rev. 2016, 116, 13279.[34] A. Rao, P. C. Y. Chow, S. Gelinas, C. W. Schlenker, C.-Z. Li, H.-L. Yip,

A. K. Y. Jen, D. S. Ginger, R. H. Friend, Nature 2013, 500, 435.[35] F. Etzold, I. A. Howard, N. Forler, A. Melnyk, D. Andrienko,

M. R. Hansen, F. Laquai, Energy Environ. Sci. 2015, 8, 1511.[36] I. Rajapaksa, K. Uenal, H. K. Wickramasinghe, Appl. Phys. Lett.

2010, 97, 073121.[37] D. Nowak, W. Morrison, H. K. Wickramasinghe, J. Jahng,

E. Potma, L. Wan, R. Ruiz, T. R. Albrecht, K. Schmidt, J. Frommer, D. P. Sanders, S. Park, Sci. Adv. 2016, 2, e1501571.

[38] J. Jahng, D. A. Fishman, S. Park, D. B. Nowak, W. A. Morrison, H. K. Wickramasinghe, E. O. Potma, Acc. Chem. Res. 2015, 48, 2671.

Documents

Side Chain Engineering on Medium Bandgap Copolymers to ...optics.nju.edu.cn/upload/20171222/201712221351484500.pdf · molecular weight sample by changing polymerization condi-tions