High Seebeck Coefficient and Power Factor in n‐Type Organic Thermoelectrics

Abstract

be attributed to the absence of an n-type equivalent of PEDOT, the overall scarcity of n-type OSCs, and the inefficiency of most n-type dopants. Since most good OSCs that are used as n-type materials like [6,6]-phenyl-C 61 -butyric acid methyl ester (PCBM) and N2200 have a lowest unoccupied molecular orbital (LUMO) around −4.0 eV, [7,8] it is very difficult to realize a stable dopant with a HOMO above −4.0 eV as would be required for efficient electron transfer. Fortunately, Wei et al. have reported a promising material, (4-(1,3-dimethyl-2,3-dihydro-1H-benzoimidazol-2-yl)phenyl) dimethylamine (N-DMBI), that can be used as stable n-type dopant due to a two-step thermally activated doping mechanism. [9] Very recently, Huang et al. used this dopant to achieve record values for PF = 105 µW m −1 K −2 and ZT = 0.11 at room temperature for the small molecule A-DCV-DPPTT. [10] Some reports also have demonstrated that dihydro-1H-benzoimidazol-2-yl (N-DBI) derivatives can improve the electrical conductivity by several orders of magnitude in n-type OTE by solution-processing. For example, Schlitz et al. have shown that solution mixtures of P(NDIOD-T2) with 9 mol% N-DBI derivatives can achieve σ ≈ 1 S m −1 and a PF over 0.6 µW m −1 K −2 . [11] Yuan et al. reported a small-molecule 2DQTT-o-OD that, with 10 wt% of a novel N-DBI derivative incorporated in solution, acquires a PF of 17.2 µW m −1 K −2 at room temperature. [12] Shi et al. achieved a high electron mobility in FBDPPV with σ ≈ 1400 S m −1 and PF ≈ 28 µW m −1 K −2 , when adding around 5 wt% N-DMBI in solution. [13] Despite the positive effect of N-DMBI and N-DBI derivatives on electrical conductivity, it probably fails to further increase σ and PF upon further increasing the volume fraction in solution mixtures due to degradation of the blend morphology. [14] Liu et al. demonstrated a modified fullerene derivative, PTEG-1, that shows an increased miscibility at the nanoscale level and thereby achieved a PF of 16.7 µW m −1 K −2 with σ ≈ 205 S m −1 at 40 mol% of N-DMBI. [15] For p-type OTE, most reported experimental data seem to follow an empirical quasi-universal power law relationship between the Seebeck coefficient S and conductivity as S ∝ σ − 1/4 . [2,3] In view of the limited number of reported data, it is still unclear whether n-type OTE also follow this power law, and different power law slopes have been reported. [15,16] Here, we introduce a novel, inverse-sequential doping procedure that mitigates the need for solvent orthogonality in conventional sequential doping to investigate the potential of deposited on a previously cast dopant 4-(1,3-dimethyl-2,3-dihydro-1H-benzoimidazol-2-yl)-N,N-diphenylaniline film to achieve a very high power factor PF ≈ 35 µW m −1 K −2 with a conductivity σ ≈ 40 S m −1 is introduced. It is also shown that n-type organic semiconductors obey the −1/4 power law relation between Seebeck coefficient S and σ that are previously found for p-type materials. An analytical model on basis of variable range hopping unifies these results. Th...

Conjugated Polymer Blends for Organic Thermoelectrics

“…and a PF close to 5 µWm -1 K -2 . [64] In line with the trend for p-type materials discussed above, the large LUMO energy difference of ∼0.67 eV between material A and B rationalizes that the maximum value of S is observed at low, ∼1 wt%, fractions of NDI-CN.…”

“…A material system for Seebeck coefficient enhancement by blending of n-type OSC has been investigated in Ref. [64] as shown in Figure 5. and a PF close to 5 µWm -1 K -2 .…”

“…On basis of the described theoretical and experimental work for both p-type and n-type OTE, DOS engineering emerges as an effective way to tune the thermoelectric properties, especially to maximize the thermopower. [56,59,60,64] Unfortunately, at the regions in parameters space where S reaches more extreme values, this comes at the cost of strongly (roughly exponentially) declining electrical conductivities, which leads to considerable decreases in power factor. [56,64] Using kinetic Monte Carlo simulations, as in panels d and e of Figure 4, it was shown in Ref.…”

“…[56,59,60,64] Unfortunately, at the regions in parameters space where S reaches more extreme values, this comes at the cost of strongly (roughly exponentially) declining electrical conductivities, which leads to considerable decreases in power factor. [56,64] Using kinetic Monte Carlo simulations, as in panels d and e of Figure 4, it was shown in Ref. [56] that for mixed systems A:B in principle the maximum PF always occurs at the pure compound that is most easily doped (A).…”

“…The thermoelectric figure of merit (ZT) is the most commonly used measure of the application potential for both p-and n-type OTE materials. The expression = 2 can be reformulated to relate the electrical conductivity to thermopower as: [55,128] = � • ,…”