2.1 Synthesis and ionic electrical modality of HAIM
We prepared the HAIM based on three criteria: (i) the iontronic host can effectively crosslink via anthracene group under UV exposure, (ii) the iontronic host must be able to store ionic charges and readily redistribute them under the external stimulations, and (iii) the iontronic host must further processing to construct hierarchical and asymmetric patterns through controllable gas phase polymerization. The first two requirements were satisfied by developing An-PIL/IL mixture with minimal free energy, among which An-PIL was used as photodimerization functional matrices to generate micropatterns and microstructures 28, and liquid IL with a substantial amount of mobile ionic charge acted as a migrant phase to realize the ionic redistribution. To satisfy the last criterion, vapor oxidative polymerization of 3,4-Ethylenedioxythiophene (EDOT) was carried out under a non-uniform chemical environment, and the resultant hierarchical morphology and asymmetrical iontronic performance came from the twice regulation of internal stress as well as distinct IL-doped PEDOT synthesized upon exposure and unexposed regions.
Figure 1A describes the overall chemical reactions and synthesized molecular structures. A random copolymer An-PIL was synthesized to bring about the photocrosslinking for iontronic host. Moreover, IL was used as a plasticizer and separable moieties to vary the glass transition temperature (\({T}_{\text{g}}\)) of An-PIL and provide microreactor for the synthesis of IL-doped PEDOT. The synthesis and characteristics of An-PIL was provided in Figure S1 and Figure S2. The fabrication process began with the preparation of An-PIL/IL film (i.e. iontronic host) comprise of different molar ratios of IL to the anthracene unit of An-PIL. Subsequently, the iontronic host became highly miscible and well dispersive with the help of the resemble structure and close characters between IL and An-PIL backbones (i.e. the solubility parameter close principle32). Thereafter, primary patterns were generated by localized photodimerization of iontronic host with photomasks and that can be understood as the self-wrinkling appeared at the exposure region due to the gradient photocrosslinking (Fig. 1B). In addition, after heating treatment at 85°C for 15 min, the IL droplets was secreted from iontronic host and firmly adhered to primary patterns. Notably, the IL droplets secretion only occurs in exposure region (i.e. primary patterns) and which can be washed out by ethanol, as shown in Figure S3. To further build hierarchical micropatterns onto the iontronic host, vapor oxidative polymerization of EDOT was conducted and resulting in secondary wrinkles formation onto previous unexposed regions. Meanwhile, the IL droplets on primary patterns acted as a microreactor to produce IL-doped PEDOT while pure PEDOT was synthesized on the surface of secondary wrinkles. As a result of formation of asymmetrical iontronic micropatterns. Figure 1C shows the optical images of a typical HAIM with a hierarchical surface containing stripe patterns (i.e. primary patterns) and random wrinkles (secondary wrinkles). The representative field-emission scanning electron microscopy (FE-SEM) image of the relevant surface morphology presented in Fig. 1D and Fig. 1E revealed that spherical IL-doped PEDOT with diameters of ~ 5 µm was aggregated onto primary patterns while flaky pure PEDOT are widely distributed onto secondary winkle.
Figure 1F demonstrates the proposed mechanisms for the ionic migration behavior of HAIM against external compressions. Owing to the presence of mobile ionic charges in IL-doped PEDOT and once HAIM were squeezed, the mobile ionic charges will be depolarized and compelled to spatially redistribute within the scaffolding polymer matrix, which we refer to as the piezoionic effect33, 34, 35. Moreover, since there is a difference between the transport capability of the ionic charges in the PEDOT/IL-rich domain in response to the built stress distribution, i.e. the cations in IL is positively charged and the interactions between cations and the main chain of PEDOT is weaker, thus, it is easier for cations to move to the opposite side away from the external force. Meanwhile, due to the stronger electrostatic force between anions (I− and TFSI−) and the positive charged PEDOT, the anions will be remained relatively near the surface. Ultimately, a negative space charge near the interface of the external force and a positive space charge at the interface far from the external force were generated, respectively. In addition, since there were no excess mobile ionic charges in pure PEDOT on secondary wrinkles and it will exhibit electrically neutral under the action of external force. Therefore, based on the distinct piezo-ionic effect of hierarchical iontronic micropatterns, the resultant HAIM exhibit asymmetrical ionic electrical performance and potential gradient under the action of an external force when connected to an external circuit.
2.2 Photodimerization and phase separation of iontronic host
To clarify the mechanism of photodimerization inducing subsequent phase separation for iontronic host, we conducted a series of controlled experiments and used the corresponding IL molar ratios-dependent micropatterns’ morphology variation to determine the factors for the initial photocrosslinking event and the resultant IL droplets secretion. The iontronic host comprise of different molar ratios of An-PIL to IL were exposed on UV light with a photomask to achieve local photodimerization, and then heated to achieve the secretion of IL droplets onto the primary patterns (Fig. 2A-C). As the molar ratios of An-PIL to IL increased, the amplitude (i.e. the height) of the resultant primary patterns was grown from 0 µm (the molar ratio of An-PIL to IL was 1:0, i.e. 0%) to 20 µm (the molar ratio of An-PIL to IL was 1:10, i.e. 91%), as shown in Fig. 2D. Subsequently, after heating the obtained primary patterns at 85°C for 15 min, IL droplets began to secrete from the surface of primary patterns when the molar ratios of An-PIL to IL was 1:1 (i.e. 50%), as shown in Figure S4. Moreover, the diameters of IL droplets were reduced from 19.7 µm to 12.9 µm until finally disappeared as the molar ratios of IL further increased. Remarkably, the as-prepared primary pattern (corresponding to the exposure region) has a clear boundary (Fig. 2B), and the IL droplets had been confined to the boundary of the primary pattern.
To gain insight into the mechanism for IL droplet secretion, we investigated the LSCM image and the corresponding phase images of a representative primary pattern in Fig. 2B. In order to better illustrate the internal phase structure change of the iontronic host, we divided the resultant surface patterns into three parts: I represent the unexposed region (i.e. the pristine iontronic host), II represent the transition region between the unexposed region and exposure region, and Ⅲ represent the exposure region (i.e. IL droplets @primary patterns), as shown in Fig. 2E-F. The LSCM image in Fig. 2F shows that evenly spaced stripe patterns were covered by IL droplets, whereas the transition region and unexposed region are flat and have no obvious change. Besides, the phase image of the unexposed region (Fig. 2G) shows a distinct nanophase-separated structure in the region I. Based on the microphase separation in the unexposed region and according to the like dissolves like rule, we proposed that the weak polarity of BA segments drives aggregation to form phase-separated BA-rich domains, on the contrary, the strong polarity of An-IL enables the An-IL-rich domains to be surrounded by IL, as shown in Fig. 2J. Meanwhile, the brighter region in the phase image (Fig. 2G) corresponds to the hard region (i.e. the An-IL moiety), and the darker region corresponds to the soft segment (i.e. the BA unit). After localized photodimerization and heating to secrete IL droplets, a relatively uniform miscible phase structure was formed in region Ⅲ (Fig. 2I). Moreover, although the transition region (II) is unexposed, however, its phase image distribution appears more uniform than that of the pristine unexposed region (I). Therefore, the mobile IL phase may migrate from the transition region to the exposure region and thus enable the An-IL and BA segment units in region II to become more homogenous.
To investigate the role of heating on IL phase separation, we heated the photodimerization cross-linked patterns (among which the molar ratios of An-PIL to IL were 1:2) at 85°C and analyzed the diameter variation of IL droplets as a function of heating time, as shown in Fig. 2K. As the heating time increased, the diameter and height of IL droplets increased from 0 µm (0 min) to 17.6 µm and 2.5 µm (12 min), respectively, and then slightly decreased until unchanged. Based on the above results, we propose that the initial mixing of An-PIL dimer and IL is thermodynamically unstable, and IL tends to phase separate and readily aggregate each other into larger droplets after heating. Moreover, since An-IL segments have good compatibility with IL with the help of the ionization of the imidazolium group and hydrogen bond interaction, therefore a miscible and stable iontronic host could be formed initially (Figure S5). Meanwhile, due to the localized photodimerization, the number of An-IL that originally interacted with IL in the exposed region became decreased, and therefore the IL droplets become freely move in the system. As the heating time increased, the free IL droplets began to collide and coalescence to grow and finally secreted out of the surface of primary patterns. Regarding the non-exposed region, the interaction between IL and An-IL was unchanged and thus there were no IL droplets separation occurred. In addition, when the molar ratios of IL reached too high, An-PIL tend to form a highly solvated network structure in IL bulk, and both An-IL and BA segments were solvated by a large amount of IL, thus the iontronic host became highly soft and the free IL droplets more likely to move into IL bulk instead of secreting out.
2.3 Ionic electrical performance of IL droplets @primary patterns
Thanks to the flexibility and adaptability of the photodimerization crosslinking and heat treatment process, a series of IL droplets @patterns (where the molar ratio of An-PIL to IL in the iontronic host is 1:2) can be readily fabricated, as shown in Fig. 3A. Due to the localized secretion of IL and thus resulting in the different content of mobile anions and cations within the unexposed and exposure regions, various micropatterns with distinct ionic conductivity can be fabricated by using corresponding photomasks and simultaneous controlling the subsequent heating time. To illustrate the unique iontronic features of IL droplets @primary patterns, the conductive AFM (C-AFM) measurement (Fig. 3B) was used to test the point contact I-V curves for different regions corresponding to a representative doughnut-like primary patterns. As shown in Figs. 3C-D, the positive and negative currents exhibit obvious nonlinear characteristics for both unexposed and exposure regions, indicating that the Schottky contact was formed between the Pt-coated tips and primary patterns. In addition, the positive bias current was significantly higher than that of the negative bias current for the unexposed region (i.e. the pristine iontronic host) by comparing the current changes in Fig. 3C. According to the ionic conduction mechanism36, 37, 38, the transport of ions among ionic polymeric networks is driven by the ion jumps over an energy barrier among coordinative polar groups on the polymer matrix. As for the iontronic host in this work, the \({T}_{\text{g}}\) decreased with the higher addition of IL (Figure S6) and thus the free ions can easily transport among An-PIL. Moreover, since the anions (free [TFSI]− groups) were bound by the polyimidazolium cation of An-PIL via the electrostatic force. Thus, the cations (free imidazolium groups) were the main mobile ions and result in a rectification effect under positive and negative bias. Besides, due to the cross-linked polymeric ionic network generated in exposure regions, free ions were hard to transport, and thus the current for exposure regions was smaller than that of the unexposed region, as shown in Fig. 3C and Fig. 3D. Further results on the current curves showed that the tunnel current for both unexposed and exposure regions were associated with IL content. The largest tunnel current appeared at the molar ratio of An-PIL to IL of 1:2, which can be explained by the intrinsic phase structure change of iontronic host. When less IL was added to An-PIL (i.e. the molar ratios of An-PIL to IL were 1:1), the rigid polymeric ionic network will hinder the ionic conduction and fewer free ions could transport above the \({T}_{\text{g}}\) of iontronic host at the same bias voltage, resulting in smaller tunnel current. In comparison, when much IL was added into An-PIL (i.e. the molar ratios of An-PIL to IL were 1:3), the iontronic host became softer and a substantial amount of IL accumulated at the bottom layer of iontronic host due to the higher density of IL than that of An-PIL. Therefore, the free ions content at the interface between the pattern and the tips was lower and resulted in a smaller tunnel current. Thus, the ionic conductivity of the patterns can be regulated by controlling the molar ratio of An-PIL to IL in iontronic hosts.
Besides, the interfacial capacitance of iontronic hosts before and after micropatterning has been investigated. Based on the polarization characteristics of An-PIL and a large number of free ions in iontronic host, the electric double layer (EDL) capacitance can readily be generated between the accumulated positive and negative charges on the two opposite electrodes and the corresponding cations and anions, as shown in Fig. 3E. The resultant interfacial capacitance against excitation frequency and molar ratios has been summarized in Fig. 2F-G. As expected, the EDL capacitance was highly frequency-dependent and the unit-area capacitance gradually decreased with a rising frequency. The capacitance-frequency curve of the exposure samples (i.e. the IL @primary patterns) exhibits a similar trend to that of unexposed samples (i.e. the pristine iontronic host) during the same frequency range (20 Hz to 2 MHz). In particular, the unit-area capacitances of pristine iontronic hosts were 1.69 µF cm− 2, 1.35 µF cm− 2, and 0.67 µF cm− 2 (20 Hz) corresponding to the molar ratios of An-PIL to IL were 1:2, 1:3, and 1:1, respectively, as shown in Fig. 3F. However, regarding the IL @primary patterns containing different molar ratios of An-PIL to IL, the unit-area capacitances declined to 0.68 µF cm− 2, 0.025 µF cm− 2, and 0.002 µF cm− 2, respectively, as shown in Fig. 3G. Thus, higher molar ratios of IL in iontronic hosts will result in an appreciable decline in the interfacial capacitance at the same excitation frequency. Finally, with the requirements of higher capacitance and highly patterned surface, the IL @primary patterns comprise of molar ratios of An-PIL to IL was 1:2 as the optimal samples for the following secondary wrinkles construction.
2.4 Construction of secondary iontronic wrinkles
Owing to the good solubility of 3,4-ethylenedioxythiophene (EDOT) in [EMIm][TFSI] (i.e. the IL in this work), the IL @primary patterns can be developed to be a microreactor vessel for the synthesis of PEDOT, as shown in Figure S7. The resultant PEDOT on primary patterns were shown in Fig. 4A-C. Among these, IL droplets firstly absorbed EDOT vapor and then processed with oxidative polymerization in the presence of iodine vapor. Thereafter, the IL-doped PEDOT was finally synthesized and secondary wrinkles appeared simultaneously, as shown in Fig. 4B. Notably, when a lower or higher molar ratio of IL was added in primary patterns (i.e., corresponding to the molar ratios of An-PIL to IL in Fig. 4a and Fig. 4c were 1:1 and 1:5, respectively), the IL droplets will not secrete onto primary patterns according to the above results in Fig. 2, and thus the EDOT vapor can only be absorbed on the surface of primary patterns, and ultimately lead to no PEDOT generated in Fig. 4A as well as the randomly distributed grainy PEDOT appeared in Fig. 4C. Subsequently, microscopic attenuated total reflection infrared spectroscopy (ATR) was used to characterize the surface components variation of IL droplets @primary patterns after vapor oxidative polymerization of PEDOT (Fig. 4D). The characteristic peaks at 1726 cm− 1 are assigned to ester groups of BA segment on An-PIL, and bands located at 1340 and 1184 cm− 1 are attributed to asymmetric and symmetric stretching vibrations of O = S = O group of TFSI−, respectively. Besides, the peak at 1049 cm− 1 was the characteristic stretching vibration of S-N-S in TFSI anions39, 40. Since the chemical structure of IL was similar to that of An-PIL, therefore no significant changes in characteristic peaks after photodimerization and heating of iontronic host, as shown in the spectra of IL @primary patterns. Moreover, new peaks at 1534, 1397, and 1250 cm− 1 were attributed to the C = C and C-C stretching vibrations of the quininoid structure of the thiophene ring, respectively. Besides, the band at about 1250 cm− 1 was due to the C-O-C bond stretching in the ethylene dioxy (alkylenedioxy) group. The series of bands agreed with the literatures41, 42, suggesting the formation of IL-doped PEDOT.
To understand the buckling mechanism for secondary wrinkles, we tracked the morphological changes of the primary patterns, as shown in Fig. 4E. As EDOT evaporated and adsorbed into IL droplets at 120°C, secondary wrinkles generated onto the unexposed regions of primary patterns after cooling. Meanwhile, the diameter of IL droplets became larger and the primary patterns turned to be swollen after absorbing EDOT. After further processing in the atmosphere of Iodine vapor, IL-doped PEDOT appeared on the surface of primary patterns, while pure PEDOT (i.e., non-IL doping PEDOT) was generated on the secondary wrinkles. To clarify the reasons for the production of secondary wrinkles, a modified bilayer system that incorporates the stiff primary patterns (surface layer) and the soft iontronic host (substrate) was proposed in Fig. 4F according to our previous work28, 43. Since the internal polymeric matrix of the primary pattern was crosslinked and showed a higher modulus, the pristine iontronic host (substrate layer) has a larger expansion coefficient relative to that of the primary patterns. The thermal expansion of the unexposed area was bound by the primary patterns during heating treatment, resulting in the accumulation of internal stress at the interface between primary patterns and unexposed regions. Moreover, EDOT vapor is more inclined to be adsorbed by IL droplets, and also a small amount of EDOT vapor will be adsorbed on unexposed regions. Therefore, the internal stress accumulated in stiff primary patterns will release during cooling and thus squeezing the softer unexposed regions to produce the typical wrinkle pattern. At the same time, IL-doped and raw PEDOT were produced on the surfaces of primary patterns and secondary wrinkles, respectively due to the oxidation of iodine vapor.
In addition, by controlling the polymerization time of PEDOT, secondary wrinkles with different amplitudes can be regulated as shown in Fig. 4G. This process can also be illustrated by a typical bilayer model incorporated with the secondary wrinkles acting as the softer substrate layer and the rigid PEDOT serving as the stiff surface layer. When the vapor polymerization time was short than 2 h, the peak of the secondary wrinkles was firstly exposed to iodine vapor and thus PEDOT was generated onto the peaks, which resulted in the interlayer stress increasing continuously, and that is the wavelength and amplitude also increasing44. When the reaction time was longer than 4 h, PEDOT will also produce in the trough region of secondary wrinkles, which causes an increase in the height of the trough, and thus the height difference between the peak and the trough decreases, and that is the wavelength and amplitude also decrease. Moreover, when the reaction time was furtherly increased to 20 h, the PEDOT layers covered on both the peaks and troughs were continuously thickened, and the wavelength and amplitude increased simultaneously. Notably, due to the excessive local stress relief, when the gas phase polymerization time was higher than 12 h, the resulting PEDOT gradually formed films and cracked, as shown in Fig. 4G. Therefore, considering the structural integrity and electrical conductivity of PEDOT, the HAMI prepared at 12 h will be used for the following study.
2.5 Asymmetric Ionic electrical performance of HAIM
Figure 5A-B showed the morphology of the as-prepared HAIM. Owing to the boundary constraints of the primary patterns, long-range wrinkle patterns in the area I parallel to the boundary were produced on the surface of the unexposed region. However, labyrinth-like wrinkle patterns in area II were generated in the unexposed regions. The distinct wrinkling patterns were caused by the different boundary constraints condition31. The LSCM images in Fig. 5C showed the amplitude of labyrinth wrinkles in area Ⅱ was lower than that of the parallel wrinkles in area Ⅰ, which was consistent with the results in Fig. 4G-H. In addition, Fig. 5D exhibited spherical PEDOT bonded to each other in area Ⅰ while flake PEDOT distributed discretely in area Ⅱ. In terms of area I (i.e., IL droplets @primary patterns), the polymerization of EDOT was carried out in IL droplets. Moreover, since the strong polarity of IL and the electrostatic interaction between IL with the PEDOT main chain, it is difficult to form a linear arrangement and stacking for the resultant spherical IL-doped PEDOT45. However, as for the chemical environment of area II, the EDOT was polymerized on the unexposed regions and without IL droplets doping, therefore the structure of the as-prepared PEDOT tended to be complete and flaky. To further illustrate the asymmetric shape of the resultant PEDOT, Energy Dispersive Spectroscopy (EDS) tests for PEDOT on both primary patterns and secondary wrinkles were conducted in Fig. 5E and Figure S8. Comparing with the change of surface atomic weight percentage for different regions, it can be found that the nitrogen content in PEDOT @secondary patterns (area Ⅱ) was close to 0% while in PEDOT @primary patterns (area I) was 3.85%. Moreover, since the EODT itself does not contain nitrogen and thus the nitrogen in polymeric EDOT can only from the contribution of IL doping. In addition, since iodide ions were the counterion for pure PEDOT in area II while [TFSI]− groups may also become the counterion of IL-doped PEODT in area I, thus the content of iodide in area II (22.54%) was obviously higher than that of area I (11.53%). In sum, the above results showed that the asymmetric PEDOT can be synthesized by developing different vapor polymerization environments on primary patterns.
The ionic electrical performance of as-prepared HAIM was performed through homemade encapsulation devices in Fig. 5F. The interface capacitance (Fig. 5G) was obtained from connecting point B and point C to a digital multi-meter. Since the IL secretion and thus the mobile ions reduced in primary patterns, the capacitance decreased from 1.69 µF cm− 2 for pristine iontronic host to that of 0.68 µF cm− 2 for IL @primary patterns, and to that of 0.007 µF cm− 2 for HAIM at the same frequency of 20 Hz. In addition, the surface resistivity (Fig. 5H) was conducted by connecting point A and point B or C to a digital multi-meter. With the production of conductive PEDOT on the surface of iontronic host, its surface resistivity decreased from the original of 3.7 MΩ•cm− 1 to 2.33 MΩ•cm− 1 for that of the PEDOT @exposure regions (area Ⅰ) and to 0.6 MΩ•cm− 1 for that of the PEDOT @unexposed region (area Ⅱ). Thereafter, the asymmetric conductive micropatterns provide sufficient conditions for the realization of piezoionic effect and resulted in a potential gradient between area Ⅰ and area Ⅱ, as shown in Fig. 5I. Generally, the piezoionic effect is a result of Donnan-like depolarization due to an inhomogeneous ionic distribution33. In terms of the HAIM, the IL-doped PEDOT initially at equilibrium has a uniform distribution of mobile species such that the electrochemical potential experienced by all species is equal and the free energy of the system is minimal. When a mechanical perturbation causes HAIM to non-homogeneously deform, the ionic species will experience a differential pressure locally and displace to result in the chemical potential change. At this depolarized state, the change in chemical potential directly corresponded to the change in electrical potential measured. This electrical potential change, \({\Delta }E\), can be can be estimated as in Eq. 1
$$\varDelta E= \frac{RT}{ZF}In\left(\frac{{\left[i\right]}_{x}}{{\left[i\right]}_{x+\varDelta x}}\right)$$
1
Where \(R\) is the gas constant, \(T\) is the temperature, \(Z\) is the valence charge of the ion, \(F\) is Faraday’s constant; \(\left[i\right]\) represents the concentration of the ith species and the subscripts represent two locations separated by a distance \(\varDelta x\) within the sample. The detailed mathematical derivation was shown in the Figure S9. Clearly, the potential gradient of HAIM is determined by the value of \({\left[i\right]}_{x}\)/\({\left[i\right]}_{x+\varDelta x}\). Higher content of ions in pristine iontronic host can easily displace to achieve a significant ions gradient, providing the larger potential gradient of -64.5 mV. However, since the diffusion and drifting of free ions in pristine iontronic host, the resultant potential gradient decreased to -1 mV after 1 min interval. Moreover, due to the photodimerization cross-linking and IL secretion among primary patterns, the IL content decreased and the ion gradient reduced, resulting in the potential gradient between the primary patterns and the unexposed region decreased to -21.6 mV and declined to -5 mV after 1 min. In particular, the potential gradient between IL-doped PEDOT on primary patterns and raw PEDOT on secondary wrinkles was stably maintained at -35.7 mV due to the stable ions gradient difference as mentioned in Fig. 1F. Thus, the vapor polymerization of asymmetric PEODT plays a vital role in the potentiometric gradient.
2.6 Regulation of the potentiometric gradient for HAIM
With the above results in mind, we further demonstrated the controlling of vapor polymerization time as the key to potentiometric gradient regulation. To facilely integrate HAIM to an external circuit, an evenly spaced stripes-like HAIM was prepared and which enabled the negative and positive to be arranged in a tandem array (Fig. 6A). The LSCM images of the as-prepared HAIM (Fig. 6B) showed a similar morphology corresponding to the above results in Fig. 5B. The micromorphology of raw PEDOT on secondary wrinkles and the IL-doped PEDOT on primary patterns were shown in Fig. 6C-F. As the polymerization times increased, spherical IL-doped PEDOT gradually joined to form a conductive film onto the primary patterns in Fig. 6F. Meanwhile, the interface capacitance between HAIM and Au-coated interdigital electrodes decreased with the polymerization time increased, as shown in Fig. 6F. Notably, the largest interfacial capacitance of HAIM appeared at the polymerization time of 12 h and which decreased slowly with frequency change. The higher interfacial capacitance indicates that more mobile ions can be polarized in IL-doped PEDOT under an external electric field and which was consistent with the surface resistivity variations in Figure H. In terms of the different surface resistivity for HAIM, both [I]− and [TFSI]− groups can become the counterion anion of PEDOT during the polymerization processes. Regrading to IL-doped PEDOT on primary patterns, the IL would have electrostatic interactions with the main chain of PEDOT and iodide ions, resulting in a reduction in the structural integrity of the as-prepared PEDOT and also a decrease in the conjugation length of π bonds46. Therefore, the surface resistivity for IL-doped PEDOT was higher than that of raw PEDOT. In addition, the surface resistivity for raw PEDOT on unexposed regions were obviously lower than that of point I (pristine iontronic host), point II (IL droplets @primary patterns), and point III (the IL @primary patterns after absorbing EDOT). Specifically, the lowest surface resistivity for raw PEDOT on secondary wrinkles dropped to 0.58 MΩ•cm− 1 when the reaction was carried out for 12 h. Meanwhile, the surface resistivity for IL-doped PEDOT on primary patterns decreased slowly with polymerization time, and the lowest value was 2.33 MΩ•cm− 1. According to these results, we proposed that the surface resistivity was initially dominated by IL content and the continuous production of PEDOT which enabled HAIM to exhibit higher conductivity. Furthermore, when the reaction time reaches 20 h, the excess iodine vapor is adsorbed on the surface of HAIM, resulting in lower ionic conduction and higher surface resistivity for HAIM.
Because IL-doped PEDOT was filled with free mobile ions, and the main chain regularity of IL-doped PEDOT was low, therefore the ions in the original equilibrium state can readily be redistributed and thus ions gradient will generate under the action of external force. In addition, the cations in IL-doped PEDOT are more easily migrated relative to the anions under the action of external force, therefore, excess anions will accumulate at the surface, resulting in a negative potential, as shown in Fig. 6I. However, IL-PEDOT will gradually form a film and become rigid with the increase of polymerization time, resulting in the reduction of ion migration and gradient under the external force, which also reduces the open circuit potential gradient between the asymmetric regions. These results indicate that the HAMI could serve as a force-induced electrical generator, implying its potential broad applications in iontronic power generators.