有机N型掺杂热电材料 2014JACS

Toward High Performance n?Type Thermoelectric Materials by Rational Modi?cation of BDPPV Backbones

Ke Shi,?Fengjiao Zhang,?Chong-An Di,*,?Tian-Wei Yan,?Ye Zou,?Xu Zhou,?Daoben Zhu,?

Jie-Yu Wang,*,?and Jian Pei*,?

?Beijing National Laboratory for Molecular Sciences,Key Laboratory of Bioorganic Chemistry and Molecular Engineering of Ministry of Education,Center for Soft Matter Science and Engineering,College of Chemistry and Molecular Engineering,Peking University, Beijing100871,China

?Beijing National Laboratory for Molecular Sciences,CAS Key Laboratory of Organic Solids,Institute of Chemistry,Chinese Academy of Sciences,Beijing100190,China

*Supporting Information

?exible electricity generation modules,organic materials have attracted increased attention in thermoelectric(TE) research.1?4The e?ciency of a TE material is determined by the dimensionless?gure of merit,ZT=S2σT/κ,where S(V K?1) is the Seebeck coe?cient or thermopower,σ(S m?1)is electrical conductivity,κ(W m?1K?1)is thermal conductivity,and T is absolute temperature.Generally,as thermal conductivities of organic materials tend to be low(0.1?1W m?1K?1)and,by contrast,electrical conductivities(10?8?104S cm?1)and Seebeck coe?cients(10?103μV K?1)range in a large scale, the TE property can also be evaluated by a parameter called power factor(PF)as PF=S2σin W m?1K?2.Viable TE devices require both electron and hole conducting materials with high power factors.Although the TE performance of p-type organic materials is rapidly advancing,5?8there are few examples of high performance n-type organic TE materials.Among these materials,vapor-doped fullerenes9,10and powder-processed organometallic poly(Ni1,1,2,2-ethenetetrathiolate)derivatives11 have demonstrated the highest n-type TE performance with electrical conductivities as high as9and40S cm?1and power factors reaching30and66μW m?1K?2,respectively.However, these materials are not amenable to solution-processing,severely restricting their extensive application.Recently,several solution-processed n-type organic TE materials were reported.Chabinyc demonstrated that solution mixtures of P(NDIOD-T2)with extrinsic dopants,N-DBI derivatives,achieved electrical conductivities of nearly10?2S cm?1and power factors of up to0.6μW m?1K?2.12Later,Segalman et al.reported that self-doped perylene diimides exhibited an electrical conductivity of 0.5S cm?1and power factors of1.4μW m?1K?2.13Nevertheless, these values are much lower than those obtained for p-type TE materials such as PEDOT,5,6which necessitates more e?ort to develop comparable n-type TE materials.

As power factor is proportional to electrical conductivityσ,the latter is the product of carrier charge q,carrier concentration n (cm?3),and carrier mobilityμ(cm2V?1s?1)asσ=nqμ. Consequently,high electron mobility and an e?cient doping process are prerequisites for high performance n-type TE materials.Although organic semiconductors with high electron mobilities have been developing rapidly,14e?cient and stable n-doping is still lagging behind and the relationship between chemical structure and doping level remains unclear.15?18In our previous work,we developed a series of BDPPV derivatives which exhibited electron mobilities up to1.70cm2V?1s?1under ambient conditions.19,20Besides,the low-lying LUMO levels (below?4.0eV)make BDPPV derivatives the most electron-de?cient conjugated polymers reported to date,which is bene?cial to an e?cient doping process.Herein,we report the electrical conductivities,Seebeck coe?cients,and the TE power factors of doped BDPPV derivatives and demonstrate that the thin?lm electrical conductivities could be dramatically enhanced through rational design of molecular structures.To investigate the relationship between chemical structure and the TE property, we develop three BDPPV derivatives,BDPPV,ClBDPPV,and FBDPPV,with varied LUMO levels and electron mobilities caused by the halogen atoms.N-DMBI is chosen as the dopant due to its good chemical stability in air and excellent doping ability for a variety of n-type semiconductors such as PCBM16 and P(NDIOD-2T).12Through simple structure modi?cation of

Received:January28,2015

Published:May21,2015

polymer backbones,we successfully modulate the doping levels to a large degree.Coupled with their excellent electron mobilities,we have achieved electrical conductivities as high as 14S cm ?1and power factors up to 28μW m ?1K ?2,which leads to the highest TE power factor that has been reported for solution processable n -type conjugated polymers.Figure 1illustrates the chemical structures of BDPPV derivatives and the dopant N-DMBI.BDPPV ,ClBDPPV ,

FBDPPV ,and N-DMBI were synthesized and fully characterized (see Supporting Information).All polymers exhibited compara-ble molecular weights (BDPPV :M n =41.8kDa,PDI =2.39;ClBDPPV :M n =38.6kDa,PDI =2.52;FBDPPV :M n =42.9kDa,PDI =2.36).The energy levels of BDPPV derivatives were explored by cyclic voltammetry to elucidate the e ?ect of halogen atoms (Table S1).After introduction of halogen atoms,both HOMO and LUMO energy levels of ClBDPPV and FBDPPV are lowered and the LUMO energy levels are more easily a ?ected.The LUMO levels of ClBDPPV and FBDPPV reach ?4.30and ?4.17eV,0.29and 0.16eV lower than that of BDPPV ,which make ClBDPPV and FBDPPV among the most electron-de ?cient conjugated polymers.Since the LUMOs of all three polymers are higher than the HOMO of N-DMBI (?4.7eV),a typical hydride transfer reaction between N-DMBI and polymers is operative.21As the feasibility of doping is necessarily accompanied by hydride or H atom transfer,the introduction of Cl and F atoms e ?ectively lowers the LUMO levels of polymers and makes ClBDPPV and FBDPPV stronger C ?H bond acceptors.With di ?erent reactivity between dopant and polymers induced by halogen atoms,the doping levels of three polymers are likely to be tuned accordingly.The optical absorption spectrum is usually utilized to indicate the doping process.Therefore,the thin ?lm absorption spectra of three BDPPV derivatives with or without doping were characterized.As shown in Figure 2b,all three undoped polymers show similar optical transitions in the visible range due to their identical conjugated backbones and no absorption over 1000nm was observed.After adding 5wt %dopant,the dark ?lms dramatically turn light and transparent (Figure 2a).Obviously,there is a considerable amount of neutral units in the ?lms that accepted electrons,leading to the formation of polarons.However,unlike their neutral counterparts,the absorption features of three doped polymers are quite distinct.In Figure 2,a new absorption band around 1100nm and extending far into the near-infrared region was observed in each doped ?lm,which reasonably originates from polaronic transitions.5,22,23Although all three doped ?lms display two major absorption bands (Band I:1000?1500nm,Band II:600?900nm),but the intensity ratios (λband I /λband II )vary di ?erently.ClBDPPV and FBDPPV have similar intensity of polaron transition,but the one of BDPPV is much lower,indicating a lower doping level under the same conditions.The absorption properties show that introduction of halogen atoms to the

polymer backbones has remarkable in ?uence on the doping extent.Further systematic study of doping levels was carried out through XPS characterizations to elucidate the chemical composition in the doped ?lms.As the doping reaction is accompanied by the generation of N-DMBI +,21the doping ratios of ?lms could be accurately estimated by analysis of the clearly

distinguishable nitrogen peaks N (1s),wherein the amide group of BDPPV derivatives (around 400eV)and the imidazole cation in N-DMBI +(around 402eV)were compared.The N (1s)electrons in the imidazole cation have a high binding energy owing to the oxidative quaternary ammonium.Besides,the N (1s)signal of neutral N-DMBI is also at 400eV,demonstrating that the newly generated N (1s)signal at 402eV is attributed to the doping reaction.The greater the area of the peak at 402eV is,the higher the doping level is.As shown in Figure 3,all three polymers under pristine conditions display the same single peak at 400eV.After being mixed with N-DMBI,however,the doping ratios among three polymers vary di ?erently.For BDPPV ,the ratio of peaks at 402and 400eV is 25:75,while,for ClBDPPV and FBDPPV ,they are both 33:67.Consequently,the XPS characterization directly con ?rms that the doping levels of ClBDPPV and FBDPPV are almost identical and are much higher than that of BDPPV ,which is in agreement with the result of optical absorption spectra.We further investigated the electronic properties of the undoped and doped ?lms of BDPPV derivatives by using ultraviolet photoemission spectros-copy (UPS)(Figure S3).All three doped ?lms exhibited

obvious

Figure 1.Chemical structures of BDPPV derivatives and n -type dopant

N-DMBI.Figure 2.(a)Photographic images of undoped and 5wt %N-DMBI doped ClBDPPV ?lms on glass substrates.(b)Thin ?lm absorption spectra of BDPPV derivatives under pristine and doping conditions (5wt %of N-DMBI)(Normalized at λmax of neutral

molecules).

Figure

3.N (1s)XPS spectra of BDPPV derivatives (a)in pristine conditions and (b)in doping conditions (5wt %of N-DMBI).Insert:N (1s)XPS of N-DMBI.

shifts of the Fermi level toward higher binding energy,equivalent to an upward movement of E F toward the LUMO level in band gap,which demonstrates the generation of free electrons and e ?ective n -doping.21To characterize the TE properties of BDPPV derivatives,the electrical conductivity and Seebeck coe ?cient were examined.Thin ?lms were prepared from solution mixtures of the polymers and N-DMBI in varied mass fraction by spin-coating on glass substrates with prepatterned gold contacts and thermal annealing at 120°C for 8h.Electrical conductivity was measured via a four-probe method,and the Seebeck coe ?cient was determined by imposing a temperature difference across the sample and measuring the thermovoltages (see Figure S4).As shown in Figure 4,the electrical conductivities of the ?lms dramatically increase by adding the dopant,and attain maxima as the mass fraction of N-DMBI in solution is 10%for BDPPV and 7%for ClBDPPV and FBDPPV .The Seebeck coe ?cients of all three polymers are negative,con ?rming that n -type electrical transport is dominant.When the electrical conductivities increase,the Seebeck coe ?cients change in opposite direction,which is in accord with their negative correlation with carrier concentrations.Further increasing the concentration of N-DMBI leads to a rapid decrease of electrical conductivities and the Seebeck coe ?cients continue to decline,possibly due to the disruption of polymer morphology or the occurrence of side reactions with excess dopants.Furthermore,Figure 4a reveals that introduction of di ?erent halogen atoms to the polymer backbones has a remarkable in ?uence on electrical conductiv-ities.The highest conductivity of BDPPV is 0.26S cm ?1,while both ClBDPPV and FBDPPV show unexpected conductivities over 4S cm ?1,higher by more than an order of magnitude.Moreover,at the optimal doping concentrations (5?10wt %),the conductivity of FBDPPV is roughly twice that of ClBDPPV .Notably,the highest conductivity of FBDPPV could reach 14S cm ?https://www.wendangku.net/doc/0a6907726.html,bination of the electrical conductivity and Seebeck coe ?cient yields a power factor as high as 28μW m ?1K ?2for FBDPPV at rt,which is the highest TE performance for solution processable n -type polymers.Besides,the dependence of conductivity and the Seebeck coe ?cient on temperature was investigated.As shown in Figure S5,the conductivity and Seebeck coe ?cient showed an increasing trend as the temper-ature increased,which is in accordance with that of typical organic thermoelectric materials reported.8The results indicated that the charge transport in these doped polymers can be described by the thermal assisted hopping mechanism.1Moreover,all doped polymers exhibited good thermal stabilities in the temperature range of 0?120°C,and their TE performance also showed good stability for at least 30days under N 2.To investigate the e ?ect of dopant amount on surface morphologies,atomic force microscopy (AFM)(Figure S6)was conducted.All undoped thin ?lms display typical ?ber-like intercalating networks.Interestingly,after being mixed with 5wt %dopant,the ?lms of all three polymers display smooth and uniform morphologies with a smaller root-mean-square (rms)roughness of 0.6nm.Unlike PCBM and P(NDIOD-T2)with high crystallinity,12,16thin ?lms of BDPPV derivatives doped by N-DMBI present no obvious particles and aggregates on the top of the surface.The low crystallinity of BDPPV derivatives ensures better dispersity of dopant in polymer ?lms,thus giving rise to a more e ?cient doping process.The smooth morphology remains almost unchanged when the dopant concentration increases to 10wt %.However,when the dopant concentration increases to 20wt %,both ?lm roughness (over 1.3nm)and crystalline size of the blend ?lms increase.Overaggregation in the blend ?lms may lower the carrier mobilities,thus leading to the decrease of electrical conductivities.The low crystallinity property of BDPPV derivatives is also supported by the grazing incident X-ray di ?raction (GIXD)(Figure S7).Three polymers all display weak out-of-plane di ?raction peaks along the q z axis,indicating the relative low crystallinity formed in thin ?lms.After doping with 5wt %N-DMBI,the weak peaks of polymers essentially remain the same,implying that the ordering of polymers is not likely to be disturbed by the dopant.To further understand the cause of drops in electrical conductivities,we measured the N (1s)XPS spectra of BDPPV derivatives under low (3wt %of N-DMBI)and high doping conditions (12wt %of N-DMBI)(Figure S8).From low (3wt %of N-DMBI)to optimized doping conditions (5wt %of N-DMBI),the ratios of peaks at 402and 400eV had a large increase,which indicated a much higher doping level.The increase of doping levels leads to a higher carrier concentration and enhanced electrical conductivity.However,from optimized (5wt %of N-DMBI)to high doping conditions (12wt %of N-DMBI),the ratios of peaks at 402and 400eV basically remained constant,which implied no further doping process.In other words,under high doping conditions (12wt %of N-DMBI),excess N-DMBI did not act as dopants but as impurities and exhibited a negative in ?uence on the polymer packing in thin ?lms.As a consequence,the decrease of electrical conductivities can be attributed to the change of ?lm morphology other than doping levels.Since ?lm morphology has limited in ?uence on Seebeck coe ?cients,the Seebeck coe ?cients remained constant as the electrical conductivities started to drop.As electrical conductivity relies on both charge carrier concentration n and carrier mobility μ,we evaluated the electron mobilities to ?gure out the di ?erence in electrical conductivity among BDPPV derivatives.Top-gate/bottom-contact (TG/BC)?eld-e ?ect transistors were fabricated for both undoped and doped polymers in varied doping concentrations (N-DMBI from 0to 5wt %).To ensure the same conditions with

electrical

Figure 4.Thermoelectric properties of doped BDPPV derivatives at di ?erent doping concentrations.(a)Electrical conductivities;(b)Seebeck coe ?cients;(c)Power factors.Each data point in Figure 4is an average of at least ?ve devices and showed good repeatability.

conductivity characterization,thin?lms were thermal annealed at 120°C for8h.As shown in Figure S9,all BDPPV derivatives based OFET devices exhibited typical n-type transporting properties with a large on?o?ratio over104under undoped conditions.When the doping percentage of N-DMBI in thin ?lms increases,the I off(measured at the zero gate voltage)of three polymers increases to di?erent degrees.For BDPPV,I off increases from10?9A to10?4A in5wt%N-DMBI doped devices.However,it is interesting to note that,for ClBDPPV and FBDPPV,I off has a much larger increase that reaches to over10?3 A with a small on?o?ratio(below2)under the same5wt%N-DMBI doped conditions.The huge di?erence in I off and on?o?ratio among the three polymers under same doping concen-trations implies di?erent doping levels in the mixed thin?lms. Further increasing the doping concentrations led to negligible dependence of the I?V characteristic on the gate voltage for all the polymers(Figure S10).In pristine conditions,the electron mobilities of BDPPV and ClBDPPV are approximately the same with an average mobility of0.5cm2V?1s?1.The mobilities of FBDPPV,by contrast,reach an average value of1.1cm2V?1s?1, which is exactly consistent with the di?erence in electrical conductivities between two polymers.Given the same doping levels of ClBDPPV and FBDPPV observed from XPS,the di?erence in their conductivities is very likely attributed to di?erent electron mobilities.The outstanding OFET perform-ance is among the highest values in n-type conjugated polymers and is bene?cial to achieving high electrical conductivity. Therefore,to obtain high electrical conductivity,both the doping level and the carrier mobility should be taken into account,neither of which is dispensable.

In conclusion,we have developed a series of high performance n-type polymers,BDPPV,ClBDPPV,and FBDPPV,for TE applications.All three polymers are capable of e?ective n-type doping after being mixed with dopant N-DMBI in solution.We demonstrate that a record conductivity of14S cm?1through a solution doping process can be obtained by rational chemical structure modi?cation.A power factor of28μW m?1K?2is also the highest value that has been reported for solution processable n-type polymers.Our investigations reveal that introduction of halogen atoms to the polymer backbones has a dramatic in?uence on not only the electron mobilities but also the doping levels,both of which are critical to the electrical conductivities. Especially,matching of energy levels of polymers and the dopant is a key issue for e?ective doping.This work opens the gate for applying the rapidly developed organic semiconductors with high carrier mobilities to the thermoelectric?eld and provides a new viewpoint to investigate the structure?property relationships of doping systems.

■ASSOCIATED CONTENT

*Supporting Information

Experimental details,synthesis,characterization,device fabrica-tion and measurements.The Supporting Information is available free of charge on the ACS Publications website at DOI:10.1021/ jacs.5b00945.

■AUTHOR INFORMATION

Corresponding Authors

*jieyuwang@https://www.wendangku.net/doc/0a6907726.html,

*jianpei@https://www.wendangku.net/doc/0a6907726.html,

*dicha@https://www.wendangku.net/doc/0a6907726.html, Notes

The authors declare no competing?nancial interest.■ACKNOWLEDGMENTS

This work was supported by the Major State Basic Research Development Program(Nos.2013CB933501and 2015CB856505)from the Ministry of Science and Technology, National Natural Science Foundation of China and Beijing Natural Science Foundation(2144049).The authors thank Prof. Ya-Pei Wang,Xin-Yue Zhang from Renmin University of China and Dr.Deng-Li Qiu from Bruker AXS GmbH Beijing Representative O?ce for AFM measurements.The authors thank beamline BL14B1(Shanghai Synchrotron Radiation Facility)for providing the beam time.

■REFERENCES

(1)Bubnova,O.;Crispin,X.Energy Environ.Sci.2012,5,9345.

(2)McGrail,B.T.;Sehirlioglu,A.;Pentzer,E.Angew.Chem.,Int.Ed. 2015,54,1710.

(3)Poehler,T.O.;Katz,H.E.Energy Environ.Sci.2012,5,8110.

(4)Zhang,Q.;Sun,Y.;Xu,W.;Zhu,D.Adv.Mater.2014,26,6829.

(5)Bubnova,O.;Khan,Z.U.;Malti,A.;Braun,S.;Fahlman,M.; Berggren,M.;Crispin,X.Nat.Mater.2011,10,429.

(6)Kim,G.H.;Shao,L.;Zhang,K.;Pipe,K.P.Nat.Mater.2013,12, 719.

(7)Bubnova,O.;Berggren,M.;Crispin,X.J.Am.Chem.Soc.2012,134, 16456.

(8)Zhang,Q.;Sun,Y.;Xu,W.;Zhu,D.Energy Environ.Sci.2012,5, 9639.

(9)Menke,T.;Ray,D.;Meiss,J.;Leo,K.;Riede,M.Appl.Phys.Lett. 2012,100,093304.

(10)Sumino,M.;Harada,K.;Ikeda,M.;Tanaka,S.;Miyazaki,K.; Adachi,C.Appl.Phys.Lett.2011,99,093308.

(11)Sun,Y.;Sheng,P.;Di,C.;Jiao,F.;Xu,W.;Qiu,D.;Zhu,D.Adv. Mater.2012,24,932.

(12)Schlitz,R.A.;Brunetti,F.G.;Glaudell,A.M.;Miller,P.L.;Brady, M.A.;Takacs,C.J.;Hawker,C.J.;Chabinyc,M.L.Adv.Mater.2014,26, 2825.

(13)Russ,B.;Robb,M.J.;Brunetti,F.G.;Miller,P.L.;Perry,E.E.; Patel,S.N.;Ho,V.;Chang,W.B.;Urban,J.J.;Chabinyc,M.L.;Hawker,

C.J.;Segalman,R.A.Adv.Mater.2014,26,3473.

(14)Zhao,Y.;Guo,Y.;Liu,Y.Adv.Mater.2013,25,5372.

(15)Naab,B.D.;Zhang,S.;Vandewal,K.;Salleo,A.;Barlow,S.; Marder,S.R.;Bao,Z.Adv.Mater.2014,26,4268.

(16)Wei,P.;Oh,J.H.;Dong,G.;Bao,Z.J.Am.Chem.Soc.2010,132, 8852.

(17)Li,C.-Z.;Chueh,C.-C.;Ding,F.;Yip,H.-L.;Liang,P.-W.;Li,X.; Jen,A.K.Y.Adv.Mater.2013,25,4425.

(18)Walzer,K.;Maennig,B.;Pfeiffer,M.;Leo,K.Chem.Rev.2007, 107,1233.

(19)Lei,T.;Xia,X.;Wang,J.-Y.;Liu,C.-J.;Pei,J.J.Am.Chem.Soc. 2014,136,2135.

(20)Lei,T.;Dou,J.-H.;Cao,X.-Y.;Wang,J.-Y.;Pei,J.J.Am.Chem.Soc. 2013,135,12168.

(21)Naab,B.D.;Guo,S.;Olthof,S.;Evans,E.G.B.;Wei,P.; Millhauser,G.L.;Kahn,A.;Barlow,S.;Marder,S.R.;Bao,Z.J.Am. Chem.Soc.2013,135,15018.

(22)Mai,C.-K.;Schlitz,R.A.;Su,G.M.;Spitzer,D.;Wang,X.;Fronk, S.L.;Cahill,D.G.;Chabinyc,M.L.;Bazan,G.C.J.Am.Chem.Soc.2014, 136,13478.

(23)Beaujuge,P.M.;Vasilyeva,S.V.;Ellinger,S.;McCarley,T.D.; Reynolds,J.R.Macromolecules2009,42,3694.