3.1. IR tunability of various doped PEDOT devices
First, we focus on the IR reflectance regulate capacity of the devices emphatically, for the extent of the ΔR variation between oxidized state and reduced state plays a crucial role during the IR modulation of the devices. Figure 2(a-h) exhibited the IR reflectance graphs of PEDOT devices doped with NaCl、 LiClO4、 H2SO4、 PTSNa、CSA、 DBSA、 PSSNa and GO in sequence under the applied potentials of + 1 V and − 2.5 V. For the voltage distribution exists between each layer of the device[34], the potential of -2.5 V is needed for the device to completely reduce PEDOT. The observed R of the PEDOT devices is actually an integration over the R values of the above PEDOT layer and the underlying Au substrate (the reflectance curve of the pure Au substrate sees Figure S1, Supporting Information). Obviously, all devices present higher reflectance as the applied voltage is -2.5 V for the PEDOT layer is substantially IR-transparent in the reduced state. Conversely, the lower reflectance of the PEDOT film in the oxidized state is observed, which is due to the generated polarons and bipolarons impedes the high reflectance of the underneath Au layer. Such behavior has also been reported by other groups[35, 36]. In addition, the reflectance modulation amplitude is low for all devices in the range of 8–14 µm. This can be attributed to all the PEDOT films with the polymerization charge of 0.3C are too thin to regulate the far IR light, as discussed detailly in our previous research[18]. Thus in the further discussion, the reflectance variation of various PEDOT devices will only concern in the 3–5 µm waveband. Figure 2(i) summarized the calculated R values of the PEDOT devices at + 1 V and − 2.5 V. Notably, the IR modulation capability of PEDOT devices doped with different dopants varies distinctly, where the PEDOT/PSSNa device possesses the optimal reflectance modulation of 0.566. Further comparison of the R values of various doped PEDOT devices in the oxidized and reduced states, it can be concluded that the differences in device reflectance management ability are mainly owing to the different reflectance of the oxidized state, while the reflectance in the reduced state is basically maintained at a high and stable level. However, the mechanism by which the PSSNa-doped PEDOT device realizes the largest IR adjustment and what the pivotal factor of different dopants influencing the variable IR reflectance remain delusive.
3.2. Morphological properties
Concerning the microstructure of the polymer films is crucial for IR regulation control[13, 37], the surface morphologies of PEDOT films doped with different dopants are studied by SEM as depicted in Fig. 3. The morphology of PEDOT doped with NaCl reveals a nanospherical aggregation with clusters of different sizes. However in the case of film doped with LiClO4, the morphology shows many tiny granular like features running across the surface. For the H2SO4 and PTSNa doped samples, they both display a dispersive net morphology generated by connecting short and thin PEDOT wires. The use of CSA and DBSA lead to porous structure, while the PEDOT layer obtained with CS− as counter ion is more porous than the obtained with DBS−. Comparatively, films prepared by PSSNa and GO are significantly smoother than those prepared by other dopants. Using PSSNa, a homogenous and flat granular morphology is obtained, consisted of highly packed grains. While the incorporation of GO and PEDOT forming a wrinkle-like sheet morphology with PEDOT grains well distributed on the GO surface. Apparently, the diversity of counter ion species intrinsically affects the structural characteristic and the morphology of PEDOT films. A smoother and flatter morphology is obtained in the presence of polyanions PSS−, which may facilitate the film having better electrochromic properties[38].
In addition, the film thickness and surface roughness are further tested by cross-sectional and AFM characterization (Fig. 4), for IR reflectance modulation is also closely related to these two factors. As reported, the IR transmittance and IR absorption of conducting polymers are greatly affected by film thickness[25, 39]. On the other hand, rough outmost surfaces can enhance interfacial interactions and trigger strong absorption[40]. From the cross-sectional Fig. 4(a-h), the thickness of various doped film with a similar polymerization charge of 0.3C is estimated to be in the range of 250–290 nm. This slight difference is not enough to cause a distinction in film reflectance variation, hence the effect of thickness can be excluded. Whereas the root mean square roughness (Rq) of PEDOT films varies distinctly on over the whole area, consistent with the phenomena observed from surface morphology as well as the cross-sectional view. The maximum Rq of PEDOT/CSA was sixteen times larger in comparison to the minimum PEDOT/PSSNa film, 110.53 nm and 7.36 nm respectively. Nevertheless, the reflectance of oxidized PEDOT/PSS is significantly lower than that of PEDOT/CSA, hence roughness is also not the decisive role affecting the film reflectance.
3.3. Study of ion transport mechanism in the electrochromic process
In general, IR electrochromic features of the conductive polymers are directly connected with their electrochemical reaction, which comprises the double transmission of electrons and ions. As the polymerization of EDOT is carried out by oxidative polymerization, the anions from the dopant are mainly migrated to the polymer chains to compensate for the charge of the polymer molecular chains to make them electrically neutral. Therefore, applying a negative potential to the PEDOT causes the film convert to a neutral reduced state resulting in an excess negative charge on the polymer backbone. At this point, ion transport occurs to maintain the electrical neutrality of the system, which includes either injecting cations into the PEDOT membrane or extracting anions from the membrane relying on the nature of the dopants. To be more specific, if the dopant is constituted of small/mobile anions, anion expel is observed to balance the negative charges on the polymer matrix. In contrast, the insertion of cation may take place if the dopant is bulky/immobile[11, 41, 42]. Hence, a deeper insight into of the ion mechanism implicated in the redox reaction of distinct doped PEDOT films should be fascinating.
To shed light on what the main mobile ion during the redox process, the XPS spectra of different doped PEDOT films at various state was systematically conducted, as shown in Fig. 5(a) and (b). Since the electrochemical process of various PEDOT films is carried out in PC/LiClO4 electrolyte, the Li and Cl signals are mainly considered. For the Li 1s spectrum (Fig. 5(a)), the Li 1s peak at 55.7 eV[43] only appeared in the reduced state of PEDOT/PSSNa and PEDOT/GO films. On the contrary, for the Cl 2p spectra (Fig. 5(b)), we could find the spin-split peaks at 207.5 eV (Cl 2p3/2) and 209.1 eV (Cl 2p1/2)[44] with a 1.6 eV separation attributed to the perchlorate characteristic in the oxidized state of PEDOT/NaCl、PEDOT/LiClO4、PEDOT/H2SO4、PEDOT/PTSNa、PEDOT/CSA and PEDOT/DBSA films. While in the reduced state, these peaks disappeared except in the film of PEDOT/LiClO4. Besides, another set of noticeable double peaks with similar intensities, ascribed to chloride, were found at 200.8 eV (Cl 2p3/2) and 202.4 eV (Cl 2p1/2)[45] in both states of PEDOT/NaCl, indicating that Cl− has been successfully doped and were not the main transport ions during the electrochemical process. In addition, neither of the Cl 2p peaks appeared in either state of PEDOT/PSS and PEDOT/GO, clarifying that the redox reactions of both films do not involve Cl.
Combined the XPS spectrum of Li 1s and Cl 2p in the oxidized and reduced state, we can clearly draw that the redox reactions of PEDOT/PSSNa and PEDOT/GO are accompanied by transport of Li+ between the polymer backbone and bulk electrolyte, while the other films are accompanied by ClO4−. Indeed, for the bulky doping ions PSS− and GO−, when electrons are added to reduce and de-dope PEDOT, it is difficult for these ions to remove from PEDOT due to the spatial site resistance effect, thus the smaller cation in the electrolyte (Li+) enter the matrix (Fig. 5(a), seventh and eighth rows, second column) and complex with doped anions. When an inverse voltage was applied, Li cation-doped anions disintegrated and Li+ migrated out from the polymer (Fig. 5(a), seventh and eighth rows, first column). This mechanism, referred to “predominant mobile cation”, is visualized in Fig. 6(a). By comparison, for the other doping ions, the electrochemical oxidation all proceeds via the emigration of electrons and interposition of ClO4− (Fig. 5(b), rows one to six, first column), while electrons are injected and ClO4− is ejected from the PEDOT film during reduction (Fig. 5(b), rows one to six, second column). This phenomenon, referred to “predominant mobile anion”, is illustrated in Fig. 6(b) and (c) to distinguish the initial doping anions of ClO4− from other small anions. Obviously, two different redox reactions are identified as follows:
$$({PEDOT)}^{{\left(m+n\right)}^{+}}:\left(m+n\right){B}^{-}+n{Li}^{+}+n{e}^{-}\left(Oxidized state\right)\leftrightarrow n\left({PEDOT}^{0}{B}^{-}{Li}^{+}\right)+m\left({PEDOT}^{+}{B}^{-}\right)\left(Reduced state\right) (1)$$
$$({PEDOT)}^{{\left(m+n\right)}^{+}}:\left(m{S}^{-}n{ClO}_{4}^{-}\right)+n{e}^{-}\left(Oxidized state\right)\leftrightarrow nPEDOT+n{ClO}_{4}^{-}+m\left({PEDOT}^{+}{S}^{-}\right)\left(Reduced state\right) (2)$$
Where B− indicates the bulky doping ions (PSS− and GO−) and S− indicates the small doping ions (Cl−、ClO4−、HSO4−、PTS−、CS− and DBS−).
To sum up, XPS experiments clearly reveals two ion transport mechanisms involved during the redox process of PEDOT films doped with different counter anions. However, considerably difference of reflectivity variations still existed despite for the same anion transport mechanism, especially for PEDOT/NaCl and PEDOT/CSA devices. This illustrates that the ion mechanism related is also not the key parameter affecting the optical properties of PEDOT films. Similar conclusions were also obtained by Petroffe et al through comparing the influence of different ionic liquids on the ion mechanism and IR electroreflective properties of PEDOT[11]. Hence, the two issues mentioned above are still blurred. In the following, we will further clarify the IR electrochromic mechanism of PEDOT by Raman、XPS、UV-Vis-NIR、FTIR characterization as well as DFT calculation.
3.4. Mechanism of the PEDOT films with variable IR reflectivity
The IR electrochromic mechanism was explored by selecting PSSNa-doped PEDOT film as an example. First, to understand the chemical structure change on a molecular level, the PEDOT/PSSNa film at different potentials are studied by Raman spectroscopy. All typical characteristic vibrations are observed for PEDOT films. In Fig. 7(a), the most intense peak at 1430 cm− 1 of neutral PEDOT is ascribed to characteristic symmetric Cα = Cβ(-O) stretching and is much sensitive to doping level. In addition, other characteristic peaks correspond to Cα = Cβ antisymmetric stretching (1522 cm− 1), Cβ-Cβ stretching (1370 cm− 1), Cα-Cα' inter-ring stretching (1271 cm− 1), C-H bending (860 cm− 1) and C-S-C deformation (699 cm− 1). The peaks at 442、573、991 cm− 1 are observed due to oxyethylene ring deformation, in agreement with Garreau et al[46].
The 1430 cm− 1 band is the most essential for it mirrors the degree of oxidation (doping) of PEDOT/PSSNa. As observed, this band expands wider and maxima is shifted towards higher wavenumber (~ 1455 cm− 1) upon oxidative doping. It should be mentioned that similar shifts were also noticed by Neprek et al. (moved from 1422 to 1436 when the SF-IL is removed from the PEDOT/PSS layer)[47] and Chiu et al. (shifted from 1414 in the dedoped PEDOT to 1445 cm− 1 in the highly doped PEDOT)[48]. These authors hypothesize that the shift results from a rise in the proportion of oxidized to neutral segments in the PEDOT chain. Moreover, Chiu et al. proposed that the strongest Cα = Cβ stretching vibration band originates from the merging of two independent bands, one caused by the symmetric Cα = Cβ stretching vibration from the neutral PEDOT components (located at 1414 cm− 1) and the other by the symmetric Cα = Cβ stretching vibration from the oxidized PEDOT components (located at 1445 cm− 1)[48]. The doping level in the PEDOT matrix affects the comprehensive strength of these bands. As the doping level increases, the composite band will be more controlled by the oxidized structure and migrate to higher wavenumbers. Therefore, the shift to higher wavenumbers indicates that the PEDOT chain is more oxidized.
Further, by curve-fitting the combined intensities of the symmetric Cα = Cβ stretching arising from the oxidized and neutral components in PEDOT/PSSNa at different potentials, the doping degree of PEDOT can be visually compared. Figure 7(b) shows the deconvolution of the Raman spectra of a PEDOT/PSSNa sample neutralized at -1 V and oxidized at + 1 V between 1300 and 1500 cm− 1. The bands of the film at various states were both deconvoluted into two vibrations at 1430 and 1455 cm− 1. Obviously, the intensities of the oxidized and neutral structures differ noticeably and the intension of the peak at 1455 cm− 1 strengthen vastly when the applied voltage is + 1 V. This demonstrates that a significant amount of delocalized polarons and bipolarons on PEDOT chains are generated at + 1 V. As the potential converted to -1 V, few polarons and bipolarons are still visible on the PEDOT chains, suggesting that a small portion of PEDOT is not participated in the redox process. In the Raman spectra of the PEDOT film dedoped with TDAE at -1 V and + 1 V (Fig. 7(c)), the band shape is essentially the same as that of -1 V. Moreover, from the curve-fitting of dedoped PEDOT film (Fig. 7(d)), it was found that the peak at 1455 cm− 1 decline sharply while still remain a small amount of strength, implying that it is extremely challenging to completely remove the dopants from PEDOT whatever via the electrochemical or chemical methods.
UV-Vis-NIR absorbance spectral was executed to further investigate the charge carrier concentration in the PEDOT/PSSNa films in different states, for neutral, polar and bipolaron states of the electrically conducting PEDOT can be distinguished through characteristic absorption features in the range of visible light to the infrared[49]. The ex situ UV-Vis-NIR absorption spectroscopy as a function of potential is indicated in Fig. 8(a). When reduced, the spectrum exhibits a distinct absorbance peak at about 500 nm assigned to π-π* transition in the PEDOT backbone[50]. Upon oxidized, this peak decreases sharply while a new absorbance at about 800 nm evolves, concurrently accompanied by increase of NIR absorption. This implies that new energy states in the band gap corresponding to polarons and bipolarons are created[35, 51]. After dedoping the PEDOT films of different states by TDAE, the UV-Vis-NIR spectra (Fig. 8(b)) displays similar shape and intensity, indicating that the two films are in the same condition after treatment. More precisely, the appeared absorption intensity at about 500 nm, the vanished absorption band at about 800 nm as well as the remarkable decline of NIR absorption, jointly implies the reduction of PEDOT and the decrease of polarons and bipolarons on the PEDOT chains.
Combined with the Raman spectra and UV-Vis-NIR spectra characterization, a significant phenomenon identified in the electrochemical redox process is the generation and decrease of polarons and bipolarons. Regarding the production of polarons and bipolarons corresponds to positive charges localization on specific thiophene rings[52], the distribution of positive charges in the polymer chains is investigated. Previous studies suggested that the oxidized form of PEDOT may possess positively polarized carbon or sulfur species within the thiophene ring[44, 45, 53]. Hence the C 1s and S 2p spectrum for PEDOT/PSSNa in the oxidized and reduced states were all obtained as shown in Fig. 9. Three main constituents are distinguished in the C 1s core-line spectra of oxidized and reduced PEDOT/PSSNa (Fig. 9(a)). The lowest binding energy is characteristic by C-C/C-H bonds (284.7 eV), accompanied by two peaks at 285.5 eV (C-S) and 286.5 eV (C-O). Besides, these strong peaks are along with a faint signal ascribed to a π-π* “shake up” owing to the aromatic rings in PSSNa as well as the thiophene rings in PEDOT at 288.8 eV[54]. Notably, some scholars also point that this peak possibly resulted from positively polarized or charged carbon[44, 55]. Whereas a comparison of oxidized and reduced PEDOT/PSSNa films shows that the C 1s core line spectra are nearly unchanged, which is contradictory to the fact that the concentration of polarons and bipolarons changes during the redox process, indicating that carbon is not the primary locus of positive charges on thiophene rings. On the other hand, the S 2p core-line spectra acquired for PEDOT/PSSNa at various states are shown in Fig. 9(b). The primary peak in each case can be matched with three constituents: neutral S (163–166 eV)、cationic S+ (164–168 eV) related with the PEDOT chains and greatly oxidized SO3− (167–171 eV) attributed to the PSS− anion[56, 57]. Every constituent is symbolized by two peaks from electrons 2p1/2 and 2p3/2, which leads to six peaks in all. Note that the intensity of the components corresponding to the cationic S+ increases obviously as the applied voltage converts from − 1 V to + 1 V, suggesting that more sulfur atoms exist in a positive circumstance resulting from the more dopants in the oxidized polymer extract charge from more thiophene units. This charge extraction and following positive charge localization on more thiophene rings unarguably leads to a greater proportion of sulfur atoms being positively polarized or partially charged[52]. The assignment of the middle two bands to the positively polarized sulfur is further favored by the truth that these constituents diminish dramatically after dedoping by TDAE, as shown in Fig. 9(c). However, a small fraction of S+ is still observed after dedoping with TDAE whatever in the oxidized or the reduced state, proving that this fraction of PEDOT always remains in the oxidized state once again. As a result, the chemical structures of the PEDOT/PSSNa at oxidized and reduced states can be recognized as shown in Fig. 9(d). More importantly, the positively polarized sulfur appears to be the symbol of polarons and bipolarons, i.e., charge carriers in the PEDOT chain. The greater of the intensity, the higher of the doping degree of PEDOT and the greater carrier concentration. Moreover, the doping level, namely the strength of charge carriers can be quantified by comparing the ratio between the S+ signals and the neutral S in the peak-fitted S 2p core-line spectra[58]. The results are listed in Table 1.
Table 1
The doping levels of the PEDOT/PSSNa films at various states.
| States | Ratio of neutral S | Ratio of S+ | Ratio of S+/neutral S |
| -1V | 0.450 | 0.115 | 0.256 |
| + 1V | 0.349 | 0.252 | 0.722 |
| -1V Dedoping | 0.679 | 0.08 | 0.118 |
| + 1V Dedoping | 0.685 | 0.064 | 0.093 |
Afterwards, the reflectance curves of PEDOT/PSSNa films in various conjugation states and the related dedoping states are inquired. As observed in Fig. 10, PEDOT is in the oxidized state and the film presents the lowest reflectance as the applied potential is + 1 V. For in this state, numerous polarons and bipolarons are formed (as shown in the lower panel of Fig. 9(b), with a doping level of 0.722), which cause strong absorption and impede IR light transmission. When the applied voltage is -1V, the PEDOT film is in the reduced state. The concentration of polarons and bipolarons decreased (related to the upper panel of Fig. 9(b), with a doping level of 0.256), thus induced the enhancement of film reflectance. After dedoping, the IR reflectance of these two states further increased, and their IR reflectance curves almost overlap. Indeed, the concentration of charge carriers all declines vastly upon dedoping (as shown in Fig. 9(c)), and the doping level in both states drops to about 0.1. Ideally, no delocalized polarons and bipolarons should exist on the reduced PEDOT chains. The reflectance curve of the PEDOT film in this state should overlap with those after being dedoped with TDAE. However, the fact is that the reflectance curve of the PEDOT film in the reduced state is significantly lower than those treated by TDAE. This implies that a tiny section of PEDOT is always in the oxidized state, causing a certain number of polarons and bipolarons always delocalized on the PEDOT backbones in the redox process, which is in accordance with the XPS analysis. Additionally, the reflectance was found to increase and decrease in-step with the elimination and formation of polarons and bipolarons delocalized on the PEDOT chains, where the increase in the oxidation level, i.e., the intensity of positively polarized sulfur supports an increscent in the concentration of charge carriers.
Finally, the influences of doping process on the structure and electronic distribution of PEDOT chains are investigated by DFT method. In place of long chain polymers, we have used oligomers of EDOT (7-mer EDOT) and SS (2-mer SS) to make huge polymeric systems tractable and save computational resources. For EDOT 7-mer, we theoretically studied the neutral EDOT(7) and complex EDOT+ 2(7)/SS− 2(2). Among them, the calculation of EDOT+ 2(7)/SS− 2(2) permitted the simulation of a bipolaron construction on the EDOT(7) chain[59]. The optimized geometries and calculated Raman spectra are shown in Fig. 11(a). It is observed that the relaxed structures of neutral EDOT(7) is planar. While if the SS− 2(2) unit is positioned above the EDOT chain, it no longer maintains complanate and bends into a concave shape to maximize the interactive with the anion. This manifests that PEDOT and PSSNa are not arrayed in a parallel manner. For the theoretical Raman spectra, the calculated DFT curves appear narrower than the experimental ones, most probably owing to the omission of film nonuniformity. In addition, the theoretical spectra show that the occurrence of SS− 2(2) in the PEDOT chain induces a slight rightward shift of the band maximum, which is consistent with the orientation of the experimental shift (see the inset, the left graph is the experiment result and the right is the calculated result). The molecular orbitals (MOs) of the EDOT(7) and complex EDOT+ 2(7)/SS− 2(2) are exhibited in Fig. 11(b). From the MOs of EDOT(7), it’s simple to note that the HOMO orbital showed π property, and the LUMO orbital possessed π* property. Thus, from HOMO to LUMO will be a π-π* transition. The calculated energy gap between HOMO and LUMO is 2.33 eV, with HOMO and LUMO energies of -3.85 and − 1.52 eV, respectively. For the EDOT+ 2(7)/SS− 2(2) structure, this value falls to 1.15 eV, with the HOMO and LUMO energies being − 4.7 and − 3.55 eV, respectively. A decline in the HOMO-LUMO gap undoubtedly suggests that the system's electronic density could be modified more easily, resulting in large carriers for the doped system with an absorption tail at the NIR region, as identified by UV-Vis-NIR spectra[51]. Beyond that, the Electrostatic potential (ESP) of the EDOT(7) and EDOT+ 2(7)/SS− 2(2) was also obtained as shown in Fig. 11(c). According to the ESP of EDOT(7), the positive and negative sections are evenly distributed on the skeleton. While this balance is broken by the incorporate of SS− 2(2). The most negative regions are found in vicinity of anions. This unbalanced ESP is in favor of charge transfer on the PEDOT chains, which is in line with the above view.
In conclusion, by integrating Raman analyses, UV-Vis-NIR spectra, XPS analysis, FTIR spectra and DFT calculation, it can be deduced that the most decisive factor for PEDOT achieving reflectance regulation lies in the generation and removal of delocalized polarons and bipolarons. Where the positively polarized sulfur appears to be the representative of polarons and bipolarons. The increase in the intensity of S+, i.e., the doping degree of PEDOT supports the greater carrier concentration. Indeed, increasing the doping level of PEDOT film is directly related to increasing the electroactivity of PEDOT electrode[60]. While higher doping level of PEDOT electrodes can provide more reactive sites on PEDOT films, thus improving the IR electrochromism performance of PEDOT films.
3.5. The charge carrier variation of PEDOT doped with different dopants
Based on the above results, the carrier concentration variation of PEDOT films doped with different dopants during redox process was further characterized by the S 2p spectrum. As displayed in Fig. 12 (a), all PEDOT films consist of six peaks. For sulfur-free doped anions such as Cl− and ClO4−, the doublet residing between 167 and 171 eV must be associated with the shake-up satellite structure, which has also been observed in other perchlorate doped sulfur-containing conjugated polymers, such as poly(2,2’-bithiophene) and poly(3-methylthiophene)[52]. While the appearance of this component in PEDOT/GO is associated with the residual sulfuric acid during the preparation of GO. In addition, a notably difference is that the intensity components corresponding to polymeric PSS counteranion is much higher than other small counteranion (167–171 eV). That is due to the higher density of sulfonate groups in PSSNa than the density of positively doped charge carried by PEDOT. Hence, while half of the sulfonate groups are employed to neutralize the doping charge, the other half cannot not be omitted from the film since they are attached to the PSSNa chain. In contrast, when it comes to small anions, only the quantity essential to neutralize the doping charge in PEDOT is taken during the electropolymerization. A common phenomenon is that all samples exhibit low S+ intensities in the reduced state and increasing intensity in the oxidized state. Nevertheless, the increase amplitude varies with the dopants. The ratio between the cationic S+ and the neutral S, which gives a direct estimation of the doping level, i.e., the charge carrier concentration, is summarized in Fig. 12(b) for different doped PEDOT films in the oxidized and reduced states (the detailed data are listed in Table S1). Meanwhile, the reflectance values of corresponding devices in both states are also summed up in Fig. 12(c). Clearly, the high ratio between the S+ and the neutral S in the oxidized state corresponds to low reflectance, and the low ratio in the reduced state corresponds to high reflectance. Moreover, it is worth note that in the same state, the trend of doping level among different doped PEDOT is opposite to the trend of reflectance of related devices, while the pace remains highly consistent. This illustrates that the difference in reflectance modulation of various anions doped PEDOT device mainly stems from the difference in doping level, i.e., charge carrier concentration. In the reduced dedoped state, the counteranion removals out or the Li+ in the electrolyte enters in the PEDOT chain, resulting in a decrease in the portion of positively sulfur atoms in the thiophene units. This small variation in dedoping levels between different anions doped PEDOT leads to the essentially same reflectance in the reduced state of according devices. While upon oxidation doping, the doping level varies considerably with doping anions, results in significant distinction in reflectance. This further reveals that the different carrier concentrations in the oxidized state is the key factor to influence the reflectance tuning of various doped PEDOT devices. The PSSNa-doped PEDOT, with the highest carrier concentration in the oxidized state, shows the lowest reflectance and the largest reflectance regulation amplitude.