Unravelling Doping Effects on PEDOT at the Molecular Level: From Geometry to Thermoelectric Transport Properties

Abstract

Tuning carrier concentration via chemical doping is the most successful strategy to optimize the thermoelectric figure of merit. Nevertheless, how the dopants affect charge transport is not completely understood. Here we unravel the doping effects by explicitly including the scattering of charge carriers with dopants on thermoelectric properties of poly(3,4-ethylenedioxythiophene), PEDOT, which is a p-type thermoelectric material with the highest figure of merit reported. We corroborate that the PEDOT exhibits a distinct transition from the aromatic to quinoid-like structure of backbone, and a semiconductor-to-metal transition with an increase in the level of doping. We identify a close-to-unity charge transfer from PEDOT to the dopant, and find that the ionized impurity scattering dominates over the acoustic phonon scattering in the doped PEDOT. By incorporating both scattering mechanisms, the doped PEDOT exhibits mobility, Seebeck coefficient and power factors in very good agreement with the experimental data, and the lightly doped PEDOT exhibits thermoelectric properties superior to the heavily doped one. We reveal that the thermoelectric transport is highly anisotropic in ordered crystals, and suggest to utilize large power factors in the direction of polymer backbone and low lattice thermal conductivity in the stacking and lamellar directions, which is viable in chain-oriented amorphous nanofibers.

Understanding morphology-mobility dependence in PEDOT:Tos

“…2(c), the mean charge carrier occupation p, i.e., the number of charge carriers per node, is a parameter that influences the mobility calculation. It is expected that it depends on the oxidation level (33% in our systems), but it should be stressed that the carrier density and the oxidation level are not the same quantities, and they can differ by orders of magnitude [20,23,26]. In our calculations we set the carrier density to 5 × 10 20 cm −3 (corresponding to p = 4 × 10 −2 ), which corresponds to the measurements of the carrier density in PEDOT reported by Park et al [58].…”

“…Indeed, so far theoretical description concerned with PEDOT and related systems utilized models not relying on its structural properties, which include cubic-grid Monte Carlo models [20,21], semi-analytical approaches [20][21][22][23] or idealized models of perfect crystal structure [24][25][26]. It is also noteworthy that Kang and Snyder [27] reported recently a general description of transport in conducting polymers that can capture essential transport parameters through a phenomenological single transport function [27].…”

Micron-thick highly conductive PEDOT films synthesized via self-inhibited polymerization: roles of anions

“…However, such PSS-based processes also experience difficulties with performance optimization. For example, the insulating PSS lamellas remain in the polymer matrix, 20,21 resulting in an amorphous polymer 22 with limited σ and S. 5,9 Compared to PEDOT:PSS, PEDOT doped with smallsized anions (S-PEDOTs), such as tosylate (OTs) or triflate (OTf), have greater potential for electrical applications because of their compact and ordered polycrystalline structure 23 that leads to higher σ and S. 22,[24][25][26] S-PEDOTs often exhibit large S 2 σ that can be further optimized by tuning the oxidation level. [27][28][29] However, these high quality films are hard to obtain due to their poor reaction controllability and difficult to put into practical use because of their limited film thickness.…”

Insulator to semimetallic transition in conducting polymers

“…Besides, this task requires approaches capable of calculating the electronic properties of systems consisting of hundreds * william.armando.munoz@liu.se † igor.zozoulenko@liu.se or thousands of atoms, which far exceeds the capabilities of standard quantum-mechanical packages. So far, the electronic properties of PEDOT and related conducting polymers have been studied on the basis of largely oversimplified models such as single polymeric chains [21][22][23][24] or infinite perfect periodic crystals [25][26][27]. None of these models are in a position to capture the morphology and the DOS of realistic PEDOT thin films.…”