Fig. 1. Schematic illustration of the fabricated vertically aligned PEDOT:PSS F-TED. (a) First-step pre-treatment; (b) second-step pre-treatment; (c) microstructure evolution after the first-step pre-treatment; (d) S2σ of the treated PEDOT:PSS after performance optimization process; (e) printing process of the TE legs arrays, followed by electron-beam evaporation to load Pt electrodes and conducting paths; (f) output power Pout and current as functions of output voltage of the as-assembled flexible TE device.

Fig. 2. (a) Measured σ of the treated PEDOT:PSS films. The x axis lists solvents used for first-step of pre-treatments; (b) evolutions of TE properties of the EG-Na2SO3-treated PEDOT:PSS films as functions of Na2SO3 concentration; (c) hall mobility (µH) and charge carrier concentration (nH) of the optimized PEDOT:PSS films as functions of Na2SO3 concentration; (d) measured S, σ and S2σ of the optimized EG-Na2SO3-treated PEDOT:PSS films as functions of measuring temperature. The inset shows the S-σ relation.

Fig. 3. GISAXS patterns of (a) PRT, (b) MTH-treated and (c) EG-treated samples; (d) the detailed line scans along the qz direction from the GISAXS patterns and (e) XPS patterns for EG-treated, MTH-treated, HSO-treated, ACT-treated, and pristine samples; (f) cross-section SEM images for (I) pristine, (II) EG-treated samples, and AFM images for (III) pristine, (IV) EG-treated samples; (g) Raman spectroscopies of EG-Na2SO3-treated samples. The caption denotes the duration of Na2SO3 concentrations.

Fig. 4. (a) Dependence of the temperature difference utilization ratio φth upon the TE legs length under various TE array distance dc. φth is defined as the ratio of the temperature difference across TE legs (ΔTTED) to available temperature difference (ΔT) between the heat sink and heat reservoir; (b) optimal leg length lop and output power density pm as functions of dc; (c) infrared image of the as-fabricated F-TED on a hotplate at 328 K. Inset is the digital image of the homemade F-TED whose internal resistance is ～9 Ω; (d) simulated gradient in output voltage Voc along the F-TED under a temperature difference of 30 K; (e) current and output power Pout as functions of Voc; (f) 240 h long-term operation test with the hot-side temperature (Th) of 328 K. The cold side was subjected to natural convection, and the room temperature was around 300 K; (g) relative electrical resistance (R/R0) and Pout of the TE module bent at various radius. Insets show the optical images of the bent TE module; h) relative electrical resistance (R/R0) and Pout stability of the TE module over 1000 cycles. Insets show the optical images of the flat and bent TE module. The bending radius is 12 mm.

Fig. 5. (a) Comparison of S2σ between other previous reports and this work [25, [38], [39], [40], [41]]; (b) Comparison of normalized output power density (N-pm) between our F-TED and the previously reported devices [13, 24, [42], [43], [44], [45]].