Electrochromism performance of electrolytic metal oxidesIn elucidating the working mechanism of EECD, as illustrated in Fig. 1b, the cathode undergoes the reversible cycling between solvated metal ions and solid metal oxide. In this regard, we selected the MnO2/Mn2+ redox pair, where MnO2 exhibits diverse advantages, including cost effectiveness, environmental friendliness, and inherent physicochemical stability. In addition, the MnSO4 aqueous solution is colorless and MnO2 displays a dark brown color, fulfilling the demands of electrochromism (Supplementary Fig. 1).Before fabricating a full device, we assessed the EC performance of the MnO2/Mn2+ reaction by analyzing the ultraviolet-visible (UV-Vis) transmittance spectra at 460 nm of a series of electrolytes containing 0.5 M MnSO4 + xM H2SO4 (x = 0, 0.3, 0.5, 1). In this acidic condition, we select fluorine-doped tin oxide (FTO) glass as a transparent conductive substrate due to its excellent stability in acidic solutions19,21. Notably, the addition of H2SO4 over 0.5 M increases the transmittance recovery from 30% to 92%. In addition, we found that increasing the H2SO4 concentration from 0.3 M to 1.0 M shortened the dissolution time from 721 s to 224 s (Supplementary Fig. 2). These findings suggest that the introduction of H2SO4 can facilitate the dissolution of MnO2. But the MnO2 deposition rate becomes more sluggish in acidic environments, as evidenced by the longer coloring time and lower deposition currents. To strike a balance, we chose 0.5 M H2SO4 + 0.5 M MnSO4 as the optimal composition for further investigation (Supplementary Figs. 3, 4).The window color is a crucial factor to consider. Notably, the deposition and dissolution of MnO2 displayed distinctive color variations ranging from light yellow to dark brown and to black when the MnO2 thickness increased from ca. 200 nm to 1000 nm (Fig. 2a). It is worth noting that the thicknesses of deposited MnO2 were determined through the cross-sectional scanning electron microscopy (SEM) (Supplementary Fig. 5). Furthermore, the energy-dispersive X-ray spectrum (EDX) results demonstrated the uniform deposition of MnO2 on the FTO electrodes (Supplementary Figs. 6, 7). In order to render the ‘perceived’ color for each optical state, we employed a spectrophotometer to qualitatively analyze the colors using three parameters (CIE 1931): L*, a*, and b*, where L* denotes lightness while a* and b* represent the ratios of green-red and blue-yellow colors, respectively (Fig. 2b and Supplementary Table 1). At the fully bleached state, the (L*, a*, b*) coordinates are (93.74, 0.04, 2.25), and they shift to (4.81, 1.34, −0.02) at the fully colored state, indicating the transition between colorless and black. To accurately identify the color change, we transformed the (L*, a*, b*) coordinate to the x*y* color space (CIE 1931), where the x-axis represents the proportion of red primary color, the y-coordinate denotes the ratio of green primary color, and the z-coordinate represents the ratio of blue primary color, which derives from the equation: x + y + z = 1.Fig. 2: Electrochromic performance of the MnO2/Mn2+ redox reaction.a Photographs of MnO2/FTO electrode, featuring color changes at different deposition/dissolution stages. b CIE L*a*b* color coordinates of MnO2/FTO electrode at various charging time. c Peak x, y color space (CIE 1931) results of MnO2/FTO electrode at different charging times (0 s to 600 s). d UV-Vis transmittance spectra of MnO2/FTO electrode charging at a potential of 1.6 V vs. standard hydrogen electrode (SHE) for 0 s, 30 s, 180 s, 300 s, and 600 s. e A plot illustrating the transmittance at 460 nm over time, with the application of deposition and dissolution potentials at 1.6 V and 0.2 V, respectively.As shown in Fig. 2c, the electrode displays a brown-yellow color when the deposition proceeds within 180 s, resulting in a selective modulation towards blue and violet light from 400 to 600 nm. Furthermore, the color of MnO2 approaches the black color at higher mass loadings, thereby expanding the adsorption lights to higher wavelengths. Therefore, the electrode transmittance continuously decreases as deposition time extends (Fig. 2d)22. Notably, the electrode achieves a low transmittance of 0–5% across the visible light range (400–800 nm) at 600 s, enabling a wide transmittance modulation range of the MnO2 EC electrode, namely a substantial optical contrast (ΔT). Figure 2e illustrates the optical response curve of MnO2 deposition/dissolution on FTO electrodes, demonstrating a high ΔT of ca. 93% between the fully colored and bleached states. It is worth noting that we selected the wavelength of 460 nm to assess the EC performance of MnO2/Mn2+. This selection is based on two main factors: the most pronounced change in transmittance observed at this wavelength and the safety concerns associated with the blue light spectrum23,24. For practical considerations, we defined the time consumed to reach 90% complete coloring and bleaching states as the coloring time (tc) and the bleaching time (tb), which are 37.7 s and 153.2 s, respectively. To assess the stability of the MnO2/Mn2+ reaction on FTO electrodes, we charged it at 1.6 V for 30 s and then discharged it at 0.2 V for 120 s. The electrode maintains a high optical contrast at 460 nm and fast response time after 2000 cycles without degradation (Supplementary Fig. 8–10).To elucidate the mechanism of MnO2 deposition and dissolution, we performed both the cyclic voltammetry (CV) test and in situ UV-Vis transmittance measurement. As shown in Fig. 3a, a single sharp anodic peak at 1.4 V corresponds to the oxidation of Mn2+ to MnO2, resulting in a rapid decrease in transmittance. X-ray diffraction (XRD) result confirmed the predominant γ-MnO2 in the deposited material (JCPDS #14-0644) (Fig. 3e, black line). In contrast, two major cathodic peaks at 1.2 V and 0.4 V, along with a minor peak at 1.0 V, revealed a complicated mechanism in MnO2 dissolution. Interestingly, the transmittance slightly changes in stage I (1.3–1.1 V) and II (1.1–0.7 V) but dramatically increases in stage III (0.7–0.2 V). Furthermore, we observed three plateaus in the galvanostatic discharging test, consistent with the CV results. The result of the electrodes prepared by galvanostatic discharging to 1.1 V in stage I, 0.7 V in stage II, and 0.4 V in stage III was analyzed (Supplementary Fig. 11). To probe the variance of chemical states, we performed the ex situ X-ray photoelectron spectroscopy (XPS) measurements for both Mn and O elements (Fig. 3b, c). The XPS O 1 s spectrum of the pristine MnO2 electrode is deconvoluted to three peaks, i.e., Mn–O–Mn at ~530.2 eV, Mn–O–H at ~531.5 eV, and H–O–H at ~533.3 eV25. In stage I, no significant change was observed in other bond peaks except for a weaker signal of Mn–O–Mn, conforming with the reduced peak intensity of Mn4+ ions in the XPS Mn 2p spectrum. These results indicate the partial dissolution of MnO2 into the electrolyte in stage I via Eq. (1)26,27:$${{{{\rm{MnO}}}}}_{2}+{4{{{\rm{H}}}}}^{+}+{2{{{\rm{e}}}}}^{-}\to {{{{\rm{Mn}}}}}^{2+}+{2{{{\rm{H}}}}}_{2}{{{\rm{O}}}}$$
(1)
Fig. 3: Deposition/dissolution chemistry of the MnO2/Mn2+ redox on FTO.a Cyclic voltammetry (CV) curve and in situ UV-Vis transmittance spectrum at 460 nm of the electrolytic MnO2/Mn2+ reaction in a 0.5 M MnSO4 + 0.5 M H2SO4 electrolyte at a scan rate of 5 mV s−1. b XPS spectra of O 1 s from deposited MnO2 electrodes to the different discharging potentials of I (1.1 V), II (0.7 V), and III (0.4 V). c XPS spectra of Mn 2p from deposited MnO2 electrodes at different discharging potentials of I (1.1 V), II (0.7 V), and III (0.4 V) d EELS spectra analyzing Mn valence for the pristine and discharged to II (0.7 V) and III (0.4 V) samples, respectively. e XRD patterns of the MnO2 electrode at the pristine state (black) and being discharged to 0.7 V (red). f HRTEM images of the as-deposited MnO2 electrode sample, and the ones being discharged to II (0.7 V) and III (0.4 V).To gain insight into the electrolytic reactions, the cathodic processes of MnO2 electrodes with different mass loadings were analyzed. Noteworthily, we found that the current responses in stage I were close in the LSV profiles, in sharp contrast to the progressively stronger peaks in stage III (Supplementary Fig. 12). Therefore, we postulated that the reduction reaction in stage I primarily happened on the surface of MnO2, where it contains more incomplete-coordinated reactive oxygen species, i.e., abundant Mn vacancies27,28,29. The instability of the surface state is conducive to the reduction of near-surface Mn4+, resulting in a higher reduction potential than that of bulk Mn4+. In both stage II and stage III, we observed more intense Mn–O–H peaks and higher ratios of Mn3+ in the O 1 s and Mn 2p spectra, suggesting the successive insertion of H+ into the oxides lattice (Fig. 3b, c). The electron energy-loss spectroscopy (EELS) revealed that the Mn–L-edge shifts to a lower oxidation energy loss at lower discharging potentials (Fig. 3d). In addition, the energy gap between L2 and L3 edges (ΔE) increased from 11.7 to 12.9 and the intensity ratio of Mn–L2,3 edge decreased from 1.35 to 2.7. These results demonstrated a constantly decreased valence state of the product upon the proton insertion. Furthermore, we conducted XRD and high-resolution transmission electron microscopy (HRTEM) to investigate the phase evolution (Fig. 3e, f). In stage II and III, a new product appeared with the XRD pattern fitted to HMnO2 (JCPDS 18-0804) while the lattice fringes of 0.46 nm corresponding to the d spaces of the (002) plane in tetragonal HMnO2. The unpronounced variations in XRD peaks relate to the slight distortion of the interplanar spacing of MnO2 during H+ insertion. The results demonstrated the intercalation of proton from the electrolyte in both stage II and III, turning Mn(IV)O2 into HMn(III)O2. Notably, the multi-step proton insertion may also relate to the complex structure of γ-MnO230,31, in which the tunnels with sizes of 2.3 Å × 4.6 Å [1 × 2] and 2.3 Å × 2.3 Å [1 × 1] are randomly distributed (Supplementary Fig. 13)32. Lastly, the HMnO2 is constantly consumed in stage III during proton insertion via Eq. (2)33.$${{{{\rm{HMnO}}}}}_{2}+{3{{{\rm{H}}}}}^{+}+{{{{\rm{e}}}}}^{-}\to {{{{\rm{Mn}}}}}^{2+}+{2{{{\rm{H}}}}}_{2}{{{\rm{O}}}}$$
(2)
which is confirmed by the significantly increased transmittance (Fig. 3a). (The schematic of dissolution mechanism of the MnO2 electrode is illustrated in Supplementary Fig. 14.)Furthermore, we found that some other metal oxides also exhibit reversible EC performance based on the deposition/dissolution chemistry, e.g., PbO2. The PbO2/Pb2+ chemistry displayed a uniform tinting and bleaching process on FTO glass and the PbO2/FTO electrode showed a similar color change to that of MnO2 (Supplementary Fig. 15). On the other hand, the PbO2/Pb2+ redox pair demonstrated some different properties from the MnO2/Mn2+ pair. For example, the deposition and dissolution of PbO2 occurs at a lower potential range between –0.4 V and 0.4 V than the MnO2 counterpart. More importantly, different from the MnO2/Mn2+ redox pair, the PbO2/Pb2+ redox pair only displays one cathodic peak in the CV curve, referring to the direct reduction of PbO2 into Pb2+. Such characteristics have a close relationship with the fast and highly reversible tinting and bleaching processes of the PbO2/Pb2+ redox (Supplementary Fig. 16). Therefore, more oxides featuring the deposition/dissolution chemistry should be explored as EC materials, where their distinctive behaviors provide opportunities for developing more advanced EECDs.Electrochromism performance of the Cu2+/Cu plating/stripping pairAs for the anode, the Cu2+/Cu redox pair was chosen due to the excellent anti-corrosion stability of Cu metal in acidic electrolytes and the high optical extinction coefficient of copper34. To maintain high light transparency, only 0.1 M CuSO4 was added into the 0.5 M MnSO4 + 0.5 M H2SO4 electrolyte. CV diagram and in situ UV-Vis transmittance spectra were analyzed on the pristine FTO electrode in a three-electrode configuration to evaluate its EC performance (Supplementary Fig. 17). It showed notable plating/stripping peaks in CV curves with a low overpotential of 0.3 V and excellent optical reversibility. XRD result demonstrated the pure phase of copper metal deposited on FTO glass (Supplementary Fig. 18). In addition, the Cu deposited on FTO electrode displayed a smooth morphology, thus facilitating the uniform coloring/bleaching of the electrode (Supplementary Fig. 19). More importantly, the Cu/FTO electrode exhibited indiscriminate light blocking across the whole testing spectrum from 400 to 800 nm, achieving a low transmittance of ca. 0% after 180 s deposition (Supplementary Fig. 20). Furthermore, the transmittance spectra of the Cu electrode in one cycle demonstrated the Cu2+/Cu redox was suitable as anode chemistry in EECDs with the advantages of fast response, good reversibility, and high optical contrast (Supplementary Fig. 21).EECDs based on dual deposition/dissolution chemistryThe concurrent electrodeposition of MnO2 on the cathode and Cu on the anode is responsible for the coloring of EECD; while in the bleaching process, the MnO2 and Cu dissolve back into the electrolyte. Linear sweep voltammetry (LSV) results demonstrated that the deposition of MnO2 and Cu occurs before the oxygen evolution reaction and hydrogen evolution reaction, promising a long cycling life and high reversibility of the EECD (Supplementary Fig. 22). To assess the electrochemical performance of the EECD, we tested CV with UV-Vis transmittance at a scan rate of 5 mV s−1 in a 0.5 M MnSO4 + 0.1 M CuSO4 + 0.5 M H2SO4 aqueous electrolyte between –0.2 V and 1.6 V vs. Cu2+/Cu. The CV profile resembles that of the MnO2 half-cell, exhibiting one anodic peak and three cathodic peaks (Supplementary Fig. 23). In addition, the coloring occurs after the anodic peak, resulting in a rapid decline of transmittance to ca. 0%, while the bleaching gradually commences after the first cathodic peak at 1.0 V and finally recovers to a transparent state after all three peaks.As shown in Fig. 4a, the EECD outperforms conventional EC devices in terms of transmittance at the bleached state of about 68-85% across the visible spectrum (400–800 nm), which is comparable to the standard low-emissivity glass window35. The performance originates from the EC-layer-free design and haze-reduction effect of liquid electrolyte. At the colored state, the EECD exhibited an impressively low transmittance of <0.01% across the visible spectrum (Fig. 4a, blue line). Furthermore, the EECD demonstrated a rapid and reversible optical response, with a short tc of 17 s and a corresponding tb of 147 s since the collective deposition of MnO2 and Cu electrodes in EECD enhances the coloring rate (Fig. 4b). Figure 4c further illustrated that EECD delivered a shorter coloring time of 60 s than that of ca. 120 s for a single MnO2 or Cu electrode to reach the 0.01% transmittance. The coloration efficiency (CE) of the EECD is about 15.3 cm2 C−1 at 460 nm (Supplementary Fig. 24). In addition, we fabricated the EC device based on MnO2/Mn2+ redox as the EC electrode, copper frame on transparent glass as the counter electrode, and the same electrolyte as a control EC device. The color change in this device is achieved solely through the deposition/dissolution of MnO2/Mn2+, and the Cu2+/Cu reaction does not contribute to the color change (Supplementary Fig. 25). As can be expected, this device delivered a stable EC performance.Fig. 4: Electrochemical and optical performance of “electrode-free” electrochromic device (EECD).a UV-Vis transmittance spectra of EECD at the colored and bleached states across the visible spectrum of 400–800 nm. b A plot of transmittance over time with applied colored and bleached voltages of 1.6 V and −0.2 V, respectively. c Transmittance-time curves of EECD and MnO2 and Cu single electrodes at 460 nm. d Thermal management performance measurement apparatus for EECD under xenon lamp illumination to simulate sunlight. The coloring device with deposited MnO2 (brown) and Cu (blue). e Real-time thermal images of EECD during illumination of xenon lamp. f Optical transmittance of the device in its transparent state. g Optical transmittance variation of EECD being colored at 1.6 V for 60 s, followed by a 7200 s power-off period, and then bleached at −0.2 V, with all measurements taken at a wavelength of 460 nm. h Optical transmittance over time curve of EECD after 10,000 cycles, measured at the wavelength of 460 nm. i Optical transmittance over time curves of EECD after being aged at 50 °C for 1 h, 25 °C for 1 h, and 0 °C for 1 h, respectively. All sample measured at the wavelength of 460 nm.The thermal management performance of a 100 cm2 square-shape EECD was further assessed by customizing a device, wherein a black rubber plate was applied as the blackbody absorber, a xenon lamp as the light source, and an IR camera was utilized to record the temperature of blackbody in real-time (Fig. 4d, e and Supplementary Fig. 26). After exposure to an illumination equivalent to 1-Sun (100 mW cm−2) for 90 s, a significantly lower temperature of 35 °C on the backside of the colored EECD was observed compared to that of over 50 °C on the backside of the bleached EECD. This distinct temperature contrast evidenced the superior capability of the colored EECD on thermal management.In the idling state, the transmittance stability of EC devices in both the bleached and colored stages, namely, bistability, is crucial to their energy efficiency. The EECD samples maintained high transmittance for over 30 days in the bleached state (Fig. 4f). More importantly, the deposited MnO2 and Cu can endure the acidic electrolyte for 2 h without transmittance degradation in the colored state (Fig. 4g and Supplementary Fig. 27). Therefore, the operation of EECD is more energy-efficient compared to many reported EC windows that require constant potential to maintain the colored state36. Cycling durability is also a key performance parameter for evaluating EC windows. After 10,000 coloring-bleaching cycles, EECD exhibits a stable EC performance with no degradation of optical contrast and only a slight decrease in response time (Fig. 4h and Supplementary Fig. 28). We postulated that the longer response time relates to the slightly increased pH value of the electrolyte after 10,000 cycles, which originates from the inevitable hydrogen evolution in the highly acidic electrolyte (Supplementary Fig. 29). The previous analysis has shown that the concentration of protons is very important for the dissolution kinetics. When there are fewer protons, it takes longer time for bleaching. In addition, EECD can exhibit proper functioning in harsh environments, i.e., 0 °C and 50 °C, which can largely expand the application scenarios, where environmental temperature requirements are not stringent (Fig. 4i). By employing differential scanning calorimetry (DSC), the electrolyte exhibited a low freezing point of −16 °C (Supplementary Fig. 30). The decreased freezing point relative to that of pure water at 0 °C is attributed to the disrupted hydrogen bonding networks of water by introducing salts and acids37. Notably, the coloring performance barely changed at the various temperatures while the bleaching time increased from 145 s to 760 s when the temperature decreased from room temperature to 0 °C. This discrepancy is dictated by the different reaction kinetics of the coloring and bleaching processes. The electrochemical impedance spectroscopy (EIS) results indicate that the dissolution kinetics of MnO2 are slower than its deposition, and the low-temperature conditions aggravate the reaction kinetics during the dissolution process (Supplementary Fig. 31).Analysis of the potential scalability of EECDsFor industrial applications, the production of large-area devices poses challenges, especially when employing intricate deposition or printing processes, which inevitably result in diminished yields and elevated production costs. In contrast, the streamlined design of EECD eliminates the need for these complex processes, offering a more efficient and practical solution. In addition, the high square resistance of the transparent electrode also presents a significant challenge in the development of large-area devices. We utilized a previously established voltage distribution equation to calculate the voltage drop (\(\Delta V\)) across a square electrode18 (Eq. (3)). The effective voltage for deposition at the center of the window is 1.57 V due to Ohmic drop (Supplementary Fig. 32). Thus, efficient switching is ensured by the voltage applied to the center part, as the deposition process begins at 1.13 V, which is the potential indicated in the cyclic voltammetry (CV) curve of the EECD (Supplementary Fig. 23).$$\Delta V=\frac{J\rho }{2t}\sqrt{\left[{\left(\frac{L}{2}\right)}^{2}-{x}^{2}\right] \left[{\left(\frac{L}{2}\right)}^{2}-{y}^{2}\right]}$$
(3)
For an EECD as large as 100 cm2 (10 cm × 10 cm) with high transmittance, uniform coloring occurs throughout the stages, making it compatible with diverse application scenarios (Fig. 5a). In addition, very little transmittance difference at both the center and edge of the windows can be observed, which confirmed the uniform tinting and bleaching of the windows (Fig. 5c). This phenomenon is quite different from the reported EC techniques, which usually show an uneven distribution of color due to the high square resistance and the unsmooth surface of TCO electrodes during coloring changing, the COMSOL Multiphysics simulation result revealed the self-regulating mechanism in EECD by modeling the electric field change during the coloring process. Particularly, with the combined effects of electromigration and diffusion, the MnO2 and Cu nuclei form on both electrodes and then continuously grow. As shown in Fig. 5b (also Supplementary Fig. 33), the modeled FTO electrode exhibited an unsmooth surface with SnO2 grains of different sizes, resulting in the inhomogeneous current density and ion flux on electrodes. The electrons and ion flux preferably concentrate on the tips of SnO2 grains38. Theoretically, the longitudinal growth of Cu into dendrites is self-accelerating, thereby short-circuiting the device39. However, the MnO2 layer helps homogenize the ion flux and electric field on both electrodes due to its low conductivity (from 10−4 to 10−3 S m−1), allowing the uniform plating of both Cu metal and MnO240. Benefiting from this mechanism, we can observe uniform tinting in an EECD window with a large area (40 cm × 40 cm), suggesting its promising scalability (Fig. 5d and Supplementary Fig. 34).Fig. 5: Scalability of “electrode-free” electrochromic device (EECD).a Photographs of a well-packaged 10 cm × 10 cm EECD sample at different coloring states (I, II, III, and IV). b COMSOL Multiphysics simulations of the dual deposition process of MnO2 and Cu from initial state to uniformly colored state (left to right). c Optical transmittance measured at the wavelength of 460 nm at the edge and center of the EC window (as marked in a) over time. d Photographs of a 40 × 40 cm2 window in the bleached state and colored state. A piece of white paper is placed at the bottom for contrast. e Estimated cost composition of commercial electrochromic devices. EECD can waive a large portion of the fabrication cost. f A radar plot comparing the major performance indexes of optical contrast (∆T), coloration time (tc), bleaching time (tb), memory effect (bistability), and cycle life. Several representative works are cited in the radar plots for comparison and the results of each characteristic are schematically illustrated semi-quantitatively.Last but not least, the EECD strategy can successfully bypass the complexities for manufacturing. This advantage extends beyond the economic spectrum; it manifests in the ease with which ordinary individuals can have these devices delivered and installed, without the need for complex and expensive manufacturing facilities. In terms of cost effectiveness, the estimated cost of conventional EC windows ranges from 180 to 250 USD m−2, in which the preparation of the EC-layer on FTO glasses, accounts for ca. 55% of the overall cost (Fig. 5e)41,42. In sharp contrast, EECD only consists of two pieces of FTO glass and a dilute electrolyte solution, resulting in a significantly reduced cost to ca. 80 USD m−2 (see estimation details in Supplementary Table 2–7). To illustrate the application of EECD smart windows, a sample was integrated with a mini-solar cell panel and a compact power management module for remote controlling by a mobile phone through Wi-Fi43 (Supplementary Fig. 35). Undoubtedly, EECD presents exceptional EC performance in various aspects, including high optical contrast, short response time, and excellent durability and bistability (Fig. 5f and Supplementary Table 8). Due to these attributes, EECDs present a robust and promising EC solution.
Electrochromic windows with fast response and wide dynamic range for visible-light modulation without traditional electrodes
