Structure and Doping Optimization of IDT-Based Copolymers for Thermoelectrics

π-conjugated backbones play a fundamental role in determining the thermoelectric (TE) properties of organic semiconductors. Understanding the relationship between the structure–property–function can help us screen valuable materials. In this study, we designed and synthesized a series of conjugated copolymers (P1, P2, and P3) based on an indacenodithiophene (IDT) building block. A copolymer (P3) with an alternating donor–acceptor (D-A) structure exhibits a narrower band gap and higher carrier mobility, which may be due to the D-A structure that helps reduce the charge carrier transport obstacles. In the end, its power factor reaches 4.91 μW m−1 K−2 at room temperature after doping, which is superior to those of non-D-A IDT-based copolymers (P1 and P2). These results indicate that moderate adjustment of the polymer backbone is an effective way to improve the TE properties of copolymers.


Introduction
Thermoelectric (TE) materials, which can directly convert thermal energy into electrical energy, provide an environmentally friendly method for energy utilization [1][2][3][4]. At present, the research directions of TE materials are mainly divided into inorganic TE and organic TE (OTE) materials [5,6]. Compared with traditional inorganic TE materials, OTE materials has good mechanical flexibility and easy to process, giving them high potential applicability for wearable and large-area applications [7][8][9]. The thermal conversion efficiency of TE materials mainly depends on the size of the dimensionless figure of merit (ZT), ZT = S 2 σT/κ, where S is the Seebeck coefficient (thermopower), σ is the electrical conductivity, T is the absolute temperature, and κ is the thermal conductivity. OTE materials usually have low κ values (0.1-0.5 W m −1 K −1 ); therefore, the study of OTE materials is mainly focused on improving the value of the power factor (S 2 σ) [7].
Proper chemical doping for conjugated polymers (CPs) is an effective way to improve charge injection [10,11], such as doped CPs that can be used as conductive interlayers for organic light-emitting diodes (OLEDs) [12] and solar cells [13]. Similarly, chemical doping in an OTE material (i.e., by introducing a molecule that oxidizes or reduces copolymers to generate a free charge carrier) is also important [14]. During the doping process, the charge carrier transfers between dopants and polymers, and the concentration or mobility of the internal carriers will influence the Seebeck coefficient and electrical conductivity of CPs [15]. Recently, there are several reports of investigating the chemical doping process to improve the TE properties of CPs [16][17][18][19] For example, Jang et al. reported the

Optical and Electrochemical Characteristics
UV-vis-NIR absorption spectra of the three polymers in the non-doped thin-film state are shown in Figure 2a. Both P1 and P2 possessed similar spectral profiles with absorption maxima at 542 nm (P1) and 511 nm (P2), while P3 exhibited two absorption peaks, the peak at 444 nm corresponds to the π-π* transition and the other peak at 625 nm should be attributed to the intramolecular charge transfer (ICT) from the electron-rich IDT segment to the electron-deficient benzothiadiazole unit [35,36]. The optical band gap (Eg opt ) values of the three polymers were calculated to be 2.02 eV (P1), 2.00 eV (P2), and 1.70 eV (P3), respectively. Cyclic voltammetry (CV) (Figures 2b) was performed to determine the highest occupied molecular orbital (HOMO) (EHOMO) and the lowest unoccupied molecular orbital (LUMO) (ELUMO) of the polymers according to the abovementioned formula. The EHOMO/ELUMO values of P1, P2, and P3 were calculated to be −5.77/−3.21 eV, −5.78/−3.28 eV, and −5.77/−3.49 eV, respectively. It is clear to see that all the three polymers exhibit

Preparation of Polymer Films
P1, P2, and P3 were dissolved in a chlorobenzene at a concentration of 5 mg mL −1 . The pristine polymer films were obtained by drop-casting the solution onto glass substrates (10 × 10 mm, washed sequentially with dichloromethane, ethanol, acetone, and isopropanol for 30 mins) under ambient conditions. Film thicknesses of the prepared samples are ranged from 6 µm to 8 µm.

Doping Experiment
FeCl 3 was dissolved in acetonitrile at a concentration of 0.1 M. The pristine polymer films were immersed in the FeCl 3 solution for different times (1,5,10,15,20, and 30 min) under ambient conditions. Finally, the residual FeCl 3 on the surface of the films was rinsed off with methanol.

Results and Discussion
P1, P2, and P3 were synthesized by Stille coupling as shown in Figure 1. The structures of the polymers were confirmed by 1 H NMR ( Figure S1). The molecular weights of three polymers were determined by gel permeation chromatography (GPC), and the number-average molecular weights (M n ) of P1, P2, and P3 were found to be 52.8, 53.7, and 58.5 kDa, respectively ( Figure S2). The thermal stability of the polymers was measured by thermal gravimetric analysis (TGA), as shown in Figure S3a, the decomposition temperature (T d , the 5% weight loss) of all polymers exceeded 380 ºC, indicating their good thermal stability. In particular, P3 exhibits the highest thermal stability, most likely due to its rigid skeleton ( Figure S3b and Table S1).

Optical and Electrochemical Characteristics
UV-vis-NIR absorption spectra of the three polymers in the non-doped thin-film state are shown in Figure 2a. Both P1 and P2 possessed similar spectral profiles with absorption maxima at 542 nm (P1) and 511 nm (P2), while P3 exhibited two absorption peaks, the peak at 444 nm corresponds to the π-π* transition and the other peak at 625 nm should be attributed to the intramolecular charge transfer (ICT) from the electron-rich IDT segment to the electron-deficient benzothiadiazole unit [35,36].  (Table 1), which are in consistent with the trend of their optical bandgaps. The difference between the band gaps obtained by the two test methods is in accordance with the trend reported in the previous literature [37]. The relatively narrower bandgap of P3 should be attributed to the charge transfer, providing the polymer to have a long effective conjugation length with more delocalized π-electrons [26].

Optical and Electrochemical Characteristics
UV-vis-NIR absorption spectra of the three polymers in the non-doped thin-film state are shown in Figure 2a. Both P1 and P2 possessed similar spectral profiles with absorption maxima at 542 nm (P1) and 511 nm (P2), while P3 exhibited two absorption peaks, the peak at 444 nm corresponds to the π-π* transition and the other peak at 625 nm should be attributed to the intramolecular charge transfer (ICT) from the electron-rich IDT segment to the electron-deficient benzothiadiazole unit [35,36]. The optical band gap (Eg opt ) values of the three polymers were calculated to be 2.02 eV (P1), 2.00 eV (P2), and 1.70 eV (P3), respectively. Cyclic voltammetry (CV) (Figures 2b) was performed to determine the highest occupied molecular orbital (HOMO) (EHOMO) and the lowest unoccupied molecular orbital (LUMO) (ELUMO) of the polymers according to the abovementioned formula. The EHOMO/ELUMO values of P1, P2, and P3 were calculated to be −5.77/−3.21 eV, −5.78/−3.28 eV, and −5.77/−3.49 eV, respectively. It is clear to see that all the three polymers exhibit relatively deep HOMO energy levels. The Eg ec values of P1, P2, and P3 were calculated to be 2.56, 2.50, and 2.28 eV, respectively (Table 1), which are in consistent with the trend of their optical bandgaps. The difference between the band gaps obtained by the two test methods is in accordance with the trend reported in the previous literature [37]. The relatively narrower bandgap of P3 should be attributed to the charge transfer, providing the polymer to have a long effective conjugation length with more delocalized π-electrons [26].  To further understand the electrochemical properties of the polymers and the distribution of delocalized electrons, we performed density functional theory (DFT) calculations on model dimers ( Figure 3). The HOMO energy of the three polymers has similar delocalization electron distribution on the backbone, which means that the introduction of acceptor into the backbone of P3 will not affect the delocalization electron distribution of HOMO energy. In contrast, the LUMO levels of P1 and P2 are delocalized on the entire polymer backbone, while the LUMO level of P3 is mainly focused on the benzothiadiazole unit with electron deficient feature. The calculated LUMO levels for the three polymers were −2.00, −2.10, and −2.70 eV, respectively, we can see that P3 showed a significantly deeper LUMO level compared to that of other two polymers, suggesting that the introduction of the D-A structure can deepen the LUMO energy to reduce the band gap (Table 1). These results are in good Polymers 2020, 12, 1463 5 of 11 agreement with the experimental values from the CV measurements (i.e., E g ec ). The side geometries of the polymers revealed the highly planar backbone features, which provides more evidence for the planar structure of three polymers that can facilitate the transfer of intermolecular charge carriers [24].  Eg ec : Electrochemical band gap. Eg opt : Optical band gap.λonset: Onset wavelength of the maximum absorption wavelength. EHOMO = −(Eox -E1/2 (Fc/Fc+) ) eV+ (−5.39 eV), ELUMO = − (Ered -E1/2 (Fc/Fc+) ) eV + (−5.39 eV), Eg ec = ELUMO -EHOMO, (Assuming that the formal potential of ferrocene is at 0.40 V versus SCE and that 0.24 V vs. NHE corresponds to 0.0 V vs SCE, then the calibration value for ferrocene is -5.39 eV) [38].
To further understand the electrochemical properties of the polymers and the distribution of delocalized electrons, we performed density functional theory (DFT) calculations on model dimers ( Figure 3). The HOMO energy of the three polymers has similar delocalization electron distribution on the backbone, which means that the introduction of acceptor into the backbone of P3 will not affect the delocalization electron distribution of HOMO energy. In contrast, the LUMO levels of P1 and P2 are delocalized on the entire polymer backbone, while the LUMO level of P3 is mainly focused on the benzothiadiazole unit with electron deficient feature. The calculated LUMO levels for the three polymers were −2.00, −2.10, and −2.70 eV, respectively, we can see that P3 showed a significantly deeper LUMO level compared to that of other two polymers, suggesting that the introduction of the D-A structure can deepen the LUMO energy to reduce the band gap (Table 1). These results are in good agreement with the experimental values from the CV measurements (i.e., Eg ec ). The side geometries of the polymers revealed the highly planar backbone features, which provides more evidence for the planar structure of three polymers that can facilitate the transfer of intermolecular charge carriers [24]. To investigate the mobilities of three polymers, we fabricated simple OFET devices without any optimizations ( Figures S4, S5), and the field effect mobility of the non-doped polymers was determined to be 3.48 × 10 −5 (P1), 13.35 × 10 −5 (P2), and 3.54 × 10 −3 cm 2 V −1 s −1 (P3). Although the mobility of P3 is lower compared to the reported values in the previous reports due to the immature OFET devices [26], P3 exhibited the highest carrier mobility among the three polymers investigated in this work, which is mainly due to its flat aromatic structure that can facilitate the intermolecular packing, thus enhancing the charge carrier transporting. Generally, CPs with higher carrier mobility is beneficial to larger electrical conductivity and higher is beneficial to larger Seebeck coefficients [39]. To investigate the mobilities of three polymers, we fabricated simple OFET devices without any optimizations ( Figures S4 and S5), and the field effect mobility of the non-doped polymers was determined to be 3.48 × 10 −5 (P1), 13.35 × 10 −5 (P2), and 3.54 × 10 −3 cm 2 V −1 s −1 (P3). Although the mobility of P3 is lower compared to the reported values in the previous reports due to the immature OFET devices [26], P3 exhibited the highest carrier mobility among the three polymers investigated in this work, which is mainly due to its flat aromatic structure that can facilitate the intermolecular packing, thus enhancing the charge carrier transporting. Generally, CPs with higher carrier mobility is beneficial to larger electrical conductivity and higher is beneficial to larger Seebeck coefficients [39].
UV-vis-NIR spectroscopy was also performed to investigate the extent of doping of the three polymers. As shown in Figure 4, after doping, new absorption peaks appeared at 872 (P1), 846 (P2), and 882 nm (P3) respectively with a broad peak at a wavelength over 1100 nm were observed, which should be generated from the formation of polarons [40,41]. As the doping time increased, the concentration of polarons were simultaneously increased.
UV-vis-NIR spectroscopy was also performed to investigate the extent of doping of the three polymers. As shown in Figure 4, after doping, new absorption peaks appeared at 872 (P1), 846 (P2), and 882 nm (P3) respectively with a broad peak at a wavelength over 1100 nm were observed, which should be generated from the formation of polarons [40,41]. As the doping time increased, the concentration of polarons were simultaneously increased.

Thin-Film Microstructure
The introduction of dopants to polymer films usually causes microscopic changes in surface morphology, resulting in different localizations of the charge carriers in the polymers films [42]. Through GI-XRD and POM ( Figure S6) analyses, we found that all three polymers are paracrystalline (see Supporting Information for details). As shown in Figure 5, the root mean square (RMS) values of the polymer films were obtained as follows: P1 (2.98 nm), P2 (2.33 nm), and P3 (1.71 nm). P3 exhibited a relatively smoother morphology compared to that of P1 and P2. After doping (immersed in 0.1 M FeCl3/acetonitrile for 15 mins), all the three polymer films showed increased RMS values (P1 (3.71 nm), P2 (2.44 nm), and P3 (4.01 nm)). It can be found that the surface roughness of P1 and P2 films only slightly increases after doping, compared to the roughness of P3 films, which indicates that the stability of the morphology of P1 and P2 films during doping is higher than that of P3 films. From the SEM ( Figure S7), we can intuitively see that the surface of the polymer film after doping is still smooth. In the EDS ( Figure S7), we also see that the residual iron element on the surface is evenly distributed without forming any aggregation, which also indicates that the FeCl3 doped the polymer uniformly.

Thin-Film Microstructure
The introduction of dopants to polymer films usually causes microscopic changes in surface morphology, resulting in different localizations of the charge carriers in the polymers films [42]. Through GI-XRD and POM ( Figure S6) analyses, we found that all three polymers are paracrystalline (see Supporting Information for details). As shown in Figure 5, the root mean square (RMS) values of the polymer films were obtained as follows: P1 (2.98 nm), P2 (2.33 nm), and P3 (1.71 nm). P3 exhibited a relatively smoother morphology compared to that of P1 and P2. After doping (immersed in 0.1 M FeCl 3 /acetonitrile for 15 mins), all the three polymer films showed increased RMS values (P1 (3.71 nm), P2 (2.44 nm), and P3 (4.01 nm)). It can be found that the surface roughness of P1 and P2 films only slightly increases after doping, compared to the roughness of P3 films, which indicates that the stability of the morphology of P1 and P2 films during doping is higher than that of P3 films. From the SEM ( Figure S7), we can intuitively see that the surface of the polymer film after doping is still smooth. In the EDS ( Figure S7), we also see that the residual iron element on the surface is evenly distributed without forming any aggregation, which also indicates that the FeCl 3 doped the polymer uniformly.
polymers. As shown in Figure 4, after doping, new absorption peaks appeared at 872 (P1), 846 (P2), and 882 nm (P3) respectively with a broad peak at a wavelength over 1100 nm were observed, which should be generated from the formation of polarons [40,41]. As the doping time increased, the concentration of polarons were simultaneously increased.

Thin-Film Microstructure
The introduction of dopants to polymer films usually causes microscopic changes in surface morphology, resulting in different localizations of the charge carriers in the polymers films [42]. Through GI-XRD and POM ( Figure S6) analyses, we found that all three polymers are paracrystalline (see Supporting Information for details). As shown in Figure 5, the root mean square (RMS) values of the polymer films were obtained as follows: P1 (2.98 nm), P2 (2.33 nm), and P3 (1.71 nm). P3 exhibited a relatively smoother morphology compared to that of P1 and P2. After doping (immersed in 0.1 M FeCl3/acetonitrile for 15 mins), all the three polymer films showed increased RMS values (P1 (3.71 nm), P2 (2.44 nm), and P3 (4.01 nm)). It can be found that the surface roughness of P1 and P2 films only slightly increases after doping, compared to the roughness of P3 films, which indicates that the stability of the morphology of P1 and P2 films during doping is higher than that of P3 films. From the SEM ( Figure S7), we can intuitively see that the surface of the polymer film after doping is still smooth. In the EDS ( Figure S7), we also see that the residual iron element on the surface is evenly distributed without forming any aggregation, which also indicates that the FeCl3 doped the polymer uniformly.

Thermoelectric Performance
To compare the different TE properties, the pristine polymer films were immersed in 0.1 M FeCl 3 /acetonitrile for different doping times. We performed the measurements for multiple times under the same condition to ensure the reproducibility of the data. As shown in Figure 6, the Seebeck coefficient (S) and electrical conductivity (σ) showed opposite trends due to the enhanced concentration of polarons. After doping the polymers for 15 mins, P1, P2, and P3 showed σ values of 0.64, 4.80, Polymers 2020, 12, 1463 7 of 11 and 9.65 S cm −1 , respectively. In the meantime, P3 exhibited the largest value of Seebeck coefficient (S) (71.26 µV K −1 ) compared to those of P1 (49.73 µV K −1 ) and P2 (54.19 µV K −1 ). Therefore, P3 showed the largest PF value of 4.91 µW m −1 K −2 . In addition, we compared the recently reported thermoelectric properties of organic matter as shown in Table S2. Obviously, it should be noted that our material still has a certain gap compared with the composite TE material, but it shows higher TE performance compared with the performance of pristine polymer TE material after doping. Indicating that IDT-based conjugated polymers are promising materials for TE applications.

Thermoelectric Performance
To compare the different TE properties, the pristine polymer films were immersed in 0.1 M FeCl3/acetonitrile for different doping times. We performed the measurements for multiple times under the same condition to ensure the reproducibility of the data. As shown in Figure 6, the Seebeck coefficient (S) and electrical conductivity (σ) showed opposite trends due to the enhanced concentration of polarons. After doping the polymers for 15 mins, P1, P2, and P3 showed σ values of 0.64, 4.80, and 9.65 S cm −1 , respectively. In the meantime, P3 exhibited the largest value of Seebeck coefficient (S) (71.26 μV K −1 ) compared to those of P1 (49.73 μV K −1 ) and P2 (54.19 μV K −1 ). Therefore, P3 showed the largest PF value of 4.91 μW m −1 K -2 . In addition, we compared the recently reported thermoelectric properties of organic matter as shown in Table S2. Obviously, it should be noted that our material still has a certain gap compared with the composite TE material, but it shows higher TE performance compared with the performance of pristine polymer TE material after doping. Indicating that IDT-based conjugated polymers are promising materials for TE applications.

Photoelectron Spectroscopy
We investigated the electronic structure and doping mechanism of polymers through X-ray photoelectron spectroscopy (XPS) and ultraviolet photoelectron spectroscopy (UPS) (Figure 7). Compared to the results from the non-doped polymer, a new C l 1s peak derived from the dopant FeCl3 was observed. More importantly, the peaks of S 2s for P1 and P2 and S 2s and N 1s for P3 ( Figure S8) shifted toward a larger binding energy caused by the electron transfer occurs between the polymers and dopant [43]. This result proves that chemical doping of the polymer by FeCl3 is mainly achieved by oxidation of the bonding atoms in the polymer. In the UPS spectra ( Figure S9), we found all the three polymers exhibited two different characteristic peaks. The peak in the range of 17.53-14.65 eV belongs to inelastic elastic electron scattering, and the peak in the range of 12.43-5.38 eV corresponds to the σ peak of the system [44]. Figure 7b shows the UPS secondary electron cut-off region of the polymer films in pristine and doped conditions. According to the formula (work function ø =21.2 − Ecutoff eV), the work functions of P1, P2, and P3 were calculated to be 4.24, 3.94, and 3.93 eV, respectively. For the secondary cutoff of the polymers, the voltage shifted to a smaller binding energy after doping, resulting in an increase in the work function (P1(4.72 eV), P2 (4.86 eV),

Photoelectron Spectroscopy
We investigated the electronic structure and doping mechanism of polymers through X-ray photoelectron spectroscopy (XPS) and ultraviolet photoelectron spectroscopy (UPS) (Figure 7). Compared to the results from the non-doped polymer, a new C l 1s peak derived from the dopant FeCl 3 was observed. More importantly, the peaks of S 2s for P1 and P2 and S 2s and N 1s for P3 ( Figure S8) shifted toward a larger binding energy caused by the electron transfer occurs between the polymers and dopant [43]. This result proves that chemical doping of the polymer by FeCl 3 is mainly achieved by oxidation of the bonding atoms in the polymer. In the UPS spectra ( Figure S9), we found all the three polymers exhibited two different characteristic peaks. The peak in the range of 17.53-14.65 eV belongs to inelastic elastic electron scattering, and the peak in the range of 12.43-5.38 eV corresponds to the σ peak of the system [44]. Figure 7b shows the UPS secondary electron cut-off region of the polymer films in pristine and doped conditions. According to the formula (work function ø = 21.2 − E cutoff eV), the work functions of P1, P2, and P3 were calculated to be 4.24, 3.94, and 3.93 eV, respectively. For the secondary cutoff of the polymers, the voltage shifted to a smaller binding energy after doping, resulting in an increase in the work function (P1(4.72 eV), P2 (4.86 eV), and P3 (4.93 eV)). Combined with the energy level structure diagram shown in Figure S10, we conclude that the Fermi level moves towards the HOMO direction when the work function increases, indicating the generation of hole carriers, and also proves that FeCl 3 is a p-type dopant for P1, P2, and P3 [35,43]. The ionization energy (IE = ø + valence band (or HOMO) onset) [45] of P1, P2, and P3 is 5.10, 5.01, and 5.18 eV, respectively ( Figure S9d); the electron affinity of FeCl 3 is 4.62 eV as documented in previous reports [11]. For p-dopable polymers, the doping efficiency should generally increase as the IE polymer − (EA) dopant difference increases, resulting in a larger thermodynamic driving force for polymer oxidation [46,47], therefore P3 with the largest HOMO onset binding energy shows the best thermoelectric performance. and P3 (4.93 eV)). Combined with the energy level structure diagram shown in Figure S10, we conclude that the Fermi level moves towards the HOMO direction when the work function increases, indicating the generation of hole carriers, and also proves that FeCl3 is a p-type dopant for P1, P2, and P3 [35,43]. The ionization energy (IE = ø + valence band (or HOMO) onset) [45] of P1, P2, and P3 is 5.10, 5.01, and 5.18 eV, respectively ( Figure S9d); the electron affinity of FeCl3 is 4.62 eV as documented in previous reports [11]. For p-dopable polymers, the doping efficiency should generally increase as the IEpolymer − (EA)dopant difference increases, resulting in a larger thermodynamic driving force for polymer oxidation [46,47], therefore P3 with the largest HOMO onset binding energy shows the best thermoelectric performance.

Conclusions
In this work, we synthesized three different IDT-based conjugated copolymers, P1, P2, and P3, to study how conjugated backbones affect the TE performances. We found that P3 with a D-A alternating backbone exhibited a narrower band gap and higher carrier mobility than those of the other two polymers. Under the same doping conditions, P3 showed the highest doping efficiency, which facilitates the charge transfer between the polymer and the dopant. As a result, P3 exhibited a maximum PF of 4.91 μW m −1 K −2 at room temperature. These results prove that the IDT-based copolymers show great prospective applications as OTE materials, and a suitable backbone structural design is important for improving the TE properties.

Conclusions
In this work, we synthesized three different IDT-based conjugated copolymers, P1, P2, and P3, to study how conjugated backbones affect the TE performances. We found that P3 with a D-A alternating backbone exhibited a narrower band gap and higher carrier mobility than those of the other two polymers. Under the same doping conditions, P3 showed the highest doping efficiency, which facilitates the charge transfer between the polymer and the dopant. As a result, P3 exhibited a maximum PF of 4.91 µW m −1 K −2 at room temperature. These results prove that the IDT-based copolymers show great prospective applications as OTE materials, and a suitable backbone structural design is important for improving the TE properties.