Polyoxometalates as promising materials for electrochromic devices

Abstract

Polyoxometalates (POMs) can act as an electron reservoir, giving rise to mixed-valence colored state species without their structural changes. This unique feature of POMs has led to the development of POM-based electrochromic devices (ECDs), which show several advantages over organic and inorganic material-based ECDs. For example, the EC properties of POMs can be tuned through molecular design. Furthermore, POMs show high chemical and UV stabilities. In this review, POM structures and the preparation methods for the EC films of POMs were comprehensively introduced. To avoid the dissolution of POM films in electrolyte media or to prepare a homogeneous transparent POM film, a stabilizer was introduced. Several types of stabilizers, including ionic types and carbon materials, were also reviewed for POM films. The EC properties of POMs depend primarily on their structural diversity. Thus, we reviewed the EC properties of different POM films prepared from diverse structures. In addition, we described the EC properties of the POM films that were obtained via different preparation methods and exhibited different morphologies. Furthermore, several types of POM-based ECDs were introduced so that not only electrochromic functions but also capacitive functions could be integrated in the ECD. This review may be of interest to researchers in the ECD field devoted to improving the material stability, optical contrast, and design of new POM structures as well as those engaged in the extension of POMs to functional materials.

---

Shi-Ming Wang

Shi-Ming Wang has been an Associate Professor at the Light Industry College in the Liaoning University since 2016. He received his Bachelor's degree from the Northeast Normal University (NENU) in 2008 and obtained his PhD degree from NENU in 2013 under the supervision of Prof. En-Bo Wang. He worked with Prof. Eunkyoung Kim as a visiting scholar (2017–2018) at the Yonsei University, Korea. His main research interest is POM-based electrochromic materials and electrochromic devices.

Jongun Hwang

Jongun Hwang is a PhD student at the Department of Chemical and Biomolecular Engineering of Yonsei University since 2017. He received his Bachelor's degree in Chemical and Biomolecular Engineering from the Yonsei University and has been working under the supervision of Prof. Eunkyoung Kim as a PhD student (2017–2019) at the Yonsei University, Korea. His main research interest is the synthesis of electronic organic materials and the design of high-performance electronic devices.

Eunkyoung Kim

Prof. Eunkyoung Kim is a Professor of the Department of Chemical and Biomolecular Engineering at Yonsei University. She received her Bachelor's degree in Chemistry from the Yonsei University, Master's degree from the Seoul National University (Seoul Korea), and PhD degree from the University of Houston, USA. Her main research interest is the development of novel methods for controlling the structure and functionality of conjugated materials for electrochromics, thermoelectrics, and photothermal switching devices.

---

1. Introduction

The electrochromic (EC) response is an interesting property of redox-active materials that show a visual color change upon electrochemical reduction or oxidation.^1–10 In particular, EC materials can contribute to the development of low-voltage working sensors, displays, and energy-saving smart windows.^5,11–17 The EC response is widely observed from transition metal oxides,^18–24 polyoxometalates (POMs),²⁵ molecular level EC materials,^26,27 and π-conjugated polymers (CPs).²⁸ In terms of the color formation at different redox states, they can be categorized as cathodically and anodically coloring EC materials. Due to their unique electro-optical properties and energy-saving potential, EC materials and devices based on transition metal oxides and π-CPs have been reviewed and updated in several studies.^16,29–37 However, reviews on polyoxometalates for EC devices are rare and need to be updated.^25,38

POMs have attracted considerable attention due to their easy synthesis and facile applicability in various fields such as in biomedicine, catalysis, and solar cells.^39–48 Among their unique properties, one of the most important properties of POMs, which gives rise to their main application, is their ability to accept electrons, yielding mixed-valence species reversibly. The POM films for EC devices are processed through low-cost electrodeposition unlike metal oxides, which require a high-vacuum process in general. Meanwhile, POMs exhibit high stability towards UV light and some of the POMs show high thermal stability over 300 °C, thus providing long-term durability for ECDs.⁴⁹ Based on these properties, POMs have been used as EC materials.²⁵

POMs consist of 3-dimensional close-packed metal oxide frameworks with a general formula of [X_xM_mO_y]ⁿ⁻(X_xM_m), where M is typically a group 6 or rarely a group 5 metal (such as Mo or W), and X is a central heteroatom from group 4 or group 5 such as P, Si, Ge, or As.^25,38 Usually, transition metals are in their high oxidation states. POMs have six basic structures: Keggin, Dawson, Anderson, Waugh, Silverton, and Lindqvist types. Among them, Keggin (x/m = 1/12) and Dawson (x/m = 2/18)-type structures are the most widely used and studied extensively.

For the Keggin structure, the size of the polyanion is around 1 nm and its lacunary derivatives (XM₁₁, XM₁₀, and XM₉) can be obtained by removing one to three {MO₆} octahedrons.^50–54 The lacunary sites can be filled with transition metals (such as V, Ni, Cu, and Co) to generate substituted Keggin structures (XM_12−nY_n, Y = V, Ni,…, n = 1–3) (Fig. 1a–g).^55,56

Fig. 1 Examples of POM structures and their size for (a) saturated Keggin, (b) mono-lacunary Keggin, (c) bi-lacunary Keggin, (d) tri-lacunary Keggin, (e) mono-substituted Keggin, (f) bi-substituted Keggin, (g) tri-substituted Keggin, (h) X₂M₁₈ (two XW₉ units), (i) saturated Dawson, (j) mono-lacunary Dawson, (k) tri-lacunary Dawson, (l) hex-lacunary Dawson, (m) mono-substituted Dawson, (n) bi-substituted Dawson, (o) tri-substituted Dawson, (p) Ln³⁺-substituted Dawson, (q) XM₆ unit (X = P, S), (r) Preyssler-type, and (s) X₈M₄₈ crown-type.

In general, the molecular size for a Dawson (∼1.5 nm)-type structure is larger than that for a Keggin-type structure. The α-Dawson-type structure X₂M₁₈O₆₄ can be regarded as the dimer of XM₉. Moreover, rotating by 60° around the axis through both heteroatoms X would afford the β-isomer.⁵⁷ The addenda atoms are partially removed or substituted by other elements to generate lacunary (P₂W₁₇, P₂W₁₅, and P₂W₁₂) or substituted Dawson structures. The general molecular formula of the substituted Dawson structures can be presented as X₂M_18−nY_nO₆₂ⁿ⁻ (X₂M_18−nY_n, X = P, Si,…; M = Mo, W; Y = V, Cr, Co,…; n = 1–3) (Fig. 1h–o).^58–60 When a rare earth metal is introduced, two mono lacunary Dawson structures become connected to one rare earth metal to form another series of substituted Dawson structures (Fig. 1p). The assembly of two or several Keggin or Dawson fragments can form other types of complex structures such as X₅M₃₀O₁₁₀ⁿ⁻ (X₅M₃₀, Preyssler-type) and P₈W₄₈O₁₈₄ (P₈W₄₈, crown-type). The X₅M₃₀O₁₁₀ⁿ⁻ structures can be regarded as the assembly of one XM₆ and two X₂M₁₂ with Na⁺ or Ln³⁺ in the center of the cage (Fig. 1l, q and r).⁶¹ The P₈W₄₈O₁₈₄ structure can be viewed as a regrouped structure of four P₂W₁₂ species (Fig. 1l and s).^58,62 There are still more diverse POMs based on Dawson-type structures, which are available in previous review papers.⁶³

POMs are a kind of inorganic EC materials; however, their preparation methods and properties are quite different from those of metal oxides. In general, POMs feature high solubility in water and in many organic solvents. Thus, POMs can be processed as a film through the solution process; however, after solvent evaporation, they form inhomogeneous films with aggregated particles. This results in very poor transparency in an ECD. Therefore, POM films for ECD are prepared simply via the electrodeposition method.

EC devices generally consist of a working electrode (WE), a counter electrode (CE), and an electrolyte layer. For a transparent electrochromic device, such as a smart window, WE and CE are transparent conductive electrodes (TCEs) such as indium tin oxide (ITO), fluorine-doped tin oxide (FTO), aluminium-doped ZnO (AZO), and a metal nanowire or mesh. The EC materials are introduced in the electrolytes as a solution mixture (Type 1) or coated on WE (Type 2, 3), as shown in Fig. 2. Typically, the soluble molecular-level organic EC materials are used as a mixture of an electrolyte (Type 1)⁶⁴ or anchored on an electrode surface (Type 2a).⁶⁵ Similarly, soluble EC polymers are used as a mixture of an electrolyte (Type 1)⁶⁴ or coated on an electrode surface (Type 2b)⁶⁶via solution casting, layer-by-layer (LBL) deposition,⁶⁷ or the spray coating method. Most CPs are introduced as thin films on WE via various deposition processes such as electrochemical polymerization and solution casting, followed by chemical polymerization.^68,69

Fig. 2 Device structure of ECDs. (a) Type 1: EC materials are used as a mixture of an electrolyte. Type 2: CE is a bare transparent conductive electrode, (b) Type 2a: EC materials are anchored on the electrode surface, (c) Type 2b: EC materials are coated on the electrode surface. Type 3: CE is coated with a functional layer, (d) Type 3a: EC materials are used for both electrodes to enhance color contrast. (e) Type 3b: a CBL is coated on CE. (f) Type 3c: EC materials are coated on WE, while CE is coated with a transparent and capacitive layer. (g) Type 3d: Both colored and bleached states are stable under the voltage-off state (V_OFF) when a HOMO-level optimized EC layer is coated on WE and an ionic liquid electrolyte is used.

The EC materials of transition metal oxides and POMs are introduced as a thin film on an electrode via several deposition methods such as thermal, electrochemical, solution casting, and LBL.^70–72 These EC materials can be coated on WE (Type 2b). In addition, they can be coated on CE along with the other types (cathodic or anodic) of EC materials on WE to enhance the EC properties (a complementary ECD, Type 3a).⁷³ The ECDs of Type 2 have a half-cell structure, while those of Type 3 exhibit a full-cell structure. Thus, Type 3 ECD can involve a charge balancing layer or a capacitive layer or both at CE (Type 3b–d). Among the ECD structures, Type 3b and 3c have recently emerged. Type 3b ECD contains a charge balancing layer (CBL) at CE,⁷⁴ so that the color contrast becomes large, while the working voltage is reduced as the charge balancing condition is reached. In Type 3c ECD, WE is coated with an EC layer, while CE is coated with a transparent and capacitive layer. Type 3c ECD combines electrochromism with electrochemical capacitors in a single ECD to afford an EC capacitive window (ECCW) or EC energy storage (EES).

Since energy loss through windows can reach to about 50% of the total energy consumption in buildings, EC materials have been developed to maintain optical memory (OM) at both colored and bleached states under the voltage-off state (V_OFF) (Type 3d).⁵ These bi-stable EC states are achieved when ECDs are under minimized interfacial electron transfer (IET) and interfacial dopant ion transfer (IDT) conditions. Bi-stable ECDs are developed by adjusting the highest occupied molecular orbital (HOMO) level of CPs; these are further used as the EC materials and an ionic liquid is used as the electrolyte of an ECD.⁵ POMs can be a good charge balancing redox shuttle to promote both bi-stability (Type 3d) as well as energy storage capacity (Type 3c). While the energy consumption to maintain their optical states is small in small-sized electronics within short-time use, it becomes significant for the operation of large-area smart windows. Thus, the combined properties of charge balancing and capacitive properties of POMs can provide new functions, such as bi-stability and energy storage capacity, for ECDs.

In this review, we provide a comprehensive introduction on the electrochromism of POMs, including the preparation method for EC films and the effect of structure and atomic composition of POMs on the EC performance. A perspective on possible research directions is also given.

2. Electrochromism in POMs

POMs can act as an electron reservoir, giving rise to mixed-valence colored state species without their structural changes.³⁹ The EC properties of POMs were first reported in 1979.⁷⁵ As EC materials, POMs have several unique advantages over other materials. From the point of view of the molecular structure, every structure has a precise content of atoms and a definitive connection mode to each atom. Although POMs are large molecules, similar to that observed for a polymer, their atomic composition and size are precisely controlled through a simple chemical synthetic process. Unlike electrochromic metal oxide materials, which may have less precisely controlled doping levels, the functional group of POMs can be stoichiometrically implemented to every unit, similar to the chemical modification of monomers for EC polymers, for high EC functions. POMs are types of inorganic EC materials, which show good UV stability similar to metal-oxide EC materials. In general, as EC materials, POMs combine the advantages of both polymer and inorganic EC materials.

The EC phenomenon in POMs occurs through the following process. The electrons at the HOMO levels are injected into the vacant orbitals involved in the oxygen-to-metal (O → M) ligand-to-metal charge transfer (LMCT) transitions after applying a more negative potential than the energy levels of the vacant orbitals. The electrons injected into the high-energy levels may also be trapped by electron traps. When the applied potential (V_ap) is more positive than the energy levels for the d¹ electron states, these electrons are returned to the electrode, yielding electrochemical oxidation of POMs. The d¹ electrons in the O → M LMCT states facilitate the absorption of visible light via intervalence charge transfer among metal centers and d–d transitions.²⁵ For example, color changes by the reduction of P₂W₁₈O₆₂⁶⁻ in an acidic electrolyte can be described by eqn (1)–(3):⁷⁶

P₂W₁₈O₆₂⁶⁻ + e⁻ = P₂W₁₈O₆₂⁷⁻ (1)

P₂W₁₈O₆₂⁷⁻ + e⁻ = P₂W₁₈O₆₂⁸⁻ (2)

P₂W₁₈O₆₂⁸⁻ + 4e⁻ + 4H⁺ = H₄P₂W₁₈O₆₂⁸⁻ (3)

Such electrochromism of POMs can be evaluated according to the properties listed below:

(1) Color contrast (ΔT, ΔA, ΔOD): the maximum transmittance, absorbance, and optical density difference between the bleached and colored states.⁷⁷

(2) Response time (t_c/t_b): the time for coloration and bleaching. It can be determined from the time for a 90% transmittance or absorbance change under an applied voltage (V_ap).

(3) Coloration efficiency (η_ce): the optical density change (ΔOD) for the charge consumed per unit of electrode area. It can be calculated from the following eqn (4):

Here, ΔOD is the optical density change at a given wavelength or λ_max, q is the injected electronic charge, S is the electrode area, ΔA is the absorbance change, and T_b and T_c are the transmittances in the bleached and colored states, respectively.

(4) Cyclic stability: the ability to retain the initial EC properties, especially the contrast and response time, when an ECD is repetitively cycled between its colored and bleached states.⁷⁸ In this review, if there is no special indication, the EC performance is reported for a three-electrode system, in which the EC film is coated on WE (ITO); also, a Pt wire is used as CE, and Ag/AgCl or a saturated calomel electrode (SCE) is used as the reference electrode in an aqueous electrolyte.

3. General methods for EC films and preparation methods for POM EC films

In general, EC films are prepared by several methods including the electrochemical method, dip-coating, spin-coating, spray coating, spray pyrolysis, chemical vapor deposition (CVD), thermal evaporation, and sputtering deposition. The electrochemical process can be classified into two categories such as electrodeposition and electropolymerization. Electrodeposition is a fast and economical method to prepare metal oxide EC films (like NiO, WO₃, and V₂O₃) on TCE and it is also suitable for large-scale manufacturing.^79–83 The electropolymerization method is usually applied for the preparation of polymer EC films.^84–86 Dip-coating, spin-coating, and spray coating methods are applied for the preparation of EC films using soluble EC polymers and metal oxides.⁸⁷ Spray pyrolysis is based on the thermal decomposition of salts, while a film is formed during the sintering process. This is a simple and relatively cost-effective method without vacuum for preparing V₂O₅-, WO₃-, and NiO-based EC films.^88–91 This method is adaptable for any shape of the substrate and is suitable for large-scale manufacturing.⁹²

The chemical vapor deposition (CVD) method is based on the chemical reaction of gaseous reactants at a heated substrate surface. PEDOT-, WO₃-, MoO₃-, and NiO-based EC films were prepared through CVD.^93–101 Thermal evaporation involves heating a solid material inside a high-vacuum chamber until it produces vapor stream that traverses the chamber and sticks to the substrate as a film. It is commonly used to produce thin WO₃, MoO₃, and NiO EC films.¹⁰² Sputtering is a physical vapor deposition method, which requires a higher vacuum condition compared with the aforementioned methods. This method has been applied to prepare metal oxide (such as WO₃, NiO, MoO₃, and V₂O₅) films.^103–106

For the preparation of POM-based EC films, a high temperature and vacuum are not used to avoid decomposition of the POM structure. Four different methods have been reported for preparing POM films, and these are tableting,⁷⁵ spin-coating, layer-by-layer (LBL),¹⁰⁷ and electrodeposition methods (Fig. 3).¹⁰⁸

Fig. 3 The illustration of the preparation methods for POM EC films. (a) Tableting method;⁷⁵ Reproduced from ref. 75, Copyright 1979, with permission from AIP Publishing. (b) Spin-coating; (c) layer-by-layer (LBL) method;¹⁰⁷ (c) Reproduced from ref. 107, Copyright 1997, with permission from AAAS. (d) Electrodeposition method.¹⁰⁸ Reproduced from ref. 108, Copyright 2013, with permission from the Royal Society of Chemistry.

In 1979, B. Tell and F. Wudl prepared a solid EC cell with H₃PW₁₂O₄₀ (PW₁₂) as an EC material. The powder-type PW₁₂ was compressed (100 kg cm⁻²) to ∼1 mm thick and ∼3 mm diameter ceramic tube on an SnO₂-coated glass (FTO). The contact of PW₁₂ and the substrate was not good, which greatly limited the performance of the EC cell (Fig. 3a). Subsequently, this method is not used to prepare a POM-based EC film anymore;⁷⁵ thus, the details of this method will not be included in this review.

POMs are quite soluble in various solvents, especially in water with water solubility up to 19 g/100 mL.¹⁰⁹ However, almost all the EC experiments are carried out in an aqueous electrolyte, which can dissolve POM films. To avoid such dissolution of the POM films in water, a stabilizer is introduced to maintain the POMs in the film state without dissolution into the electrolyte. Furthermore, POMs easily crystallize and aggregate in a film state to result in a hazy film. Thus, a stabilizer is added to form a transparent film without haziness. These stabilizations and good dispersion of POMs in solid-state films are crucial for high EC performances. Therefore, ionic and polar stabilizers including large organic or polymeric cations, which induce cation–anion interaction, chemical bonding, or hydrogen bond interactions with POMs, are introduced in the POM solution. Conjugated polymers, carbon materials, and metal oxide-based semiconductors have been used as stabilizers to make a POM film through a solution and wet chemistry process for an ECD.

3.1 Spin-coating method

Spin-coating is an effective and low-cost method to prepare a homogeneous thin film and it has been used for POM-based EC films (Fig. 3b). Generally, if the EC films are examined in an aqueous electrolyte, the stabilizer and POMs should be mixed first and then, the mixture is spin-coated. If there is no stabilizer for POMs, the electrolyte should be in the solid state to avoid the dissolution of POMs.

In 1994, Sanae et al. reported a POM-based EC film prepared by spin-coating.¹¹⁰ In 1997, a solution of [(CH₃)₂NC₆H₄NH(CH₃)₂]₄[(C₄H₉)₄N]SiMo₁₂O₄₀ in DMF was spin-coated to prepare an EC film, which showed color change from yellow to green and violet under the potential change from 1.0 V to 0 V and −1.0 V, respectively, (Fig. 4) in a 3-electrode cell.¹¹¹ In the same year, an H₅PMo₁₀V₂O₄₀ film was prepared without any stabilizer and was applied to a solid-state EC window; it consisted of Li⁺-doped oxymethylene polyoxyethylene (OMPE) as the solid polymer electrolyte and Prussian blue-coated ITO as CE. A change in absorbance from 0.73 (−1.5 V) to 0.48 (1.0 V) was detected at 700 nm.¹¹²

Fig. 4 (a) Molecular structure of the complex; (b) the illustration of the configuration for cuvette cell; (c) absorption spectra of complex-based EC film at ① 1.0, ② 0, and ③ −1.0 V.¹¹² Reproduced from ref. 112, Copyright 1997, with permission from American Chemical Society.

In 2003, alkylammonium was used as a stabilizer to prepare alkylammonium molybdates (AAMs) (C₁₈H₃₇NH₃)₅HMo₇O₂₄ by spin-coating. AAMs showed a laminated super-lattice structure with a d-spacing of 3.45 nm and exhibited color changes under a negative potential.¹¹³

Although the spin-coating method for the preparation of POM films was considered until 2003, the EC response time, ΔT, and η_ce of POM EC films prepared by the spin-coating method have not been reported. Since then, no reports have been found on the POM films for ECD prepared by the spin-coating method. This can be because the POM films obtained from solution-coating methods afford rather low ΔA (Fig. 4c) and POMs are easily aggregated after the removal of solvents, resulting in a hazy film.

3.2 Layer-by-layer method

In 1997, Gero Decher developed the LBL method, which can be easily carried out by the consecutive adsorption of polyanions and polycations.^107,114 As shown in Fig. 3c, steps 1 and 3 represent the adsorptions of a polyanion and polycation, respectively; steps 2 and 4 are washing steps. The four steps form the basic building sequence for the simplest film architecture (A/B)_n.¹⁰⁷

The LBL method began to spread rapidly and nearly all of the POM-based films were fabricated by using this method.^115–119 An EC film prepared by LBL shows high transparency at the bleaching state and cyclic stability. In the LBL method, to prepare the POM EC film, commonly used polycations are PAH: poly(allylamine hydrochloride), PSS: poly(styrene sulfonate), PDDA: poly(diallyl dimethyl ammonium chloride), and PEI: poly(ethylenimine).

3.3 Electrodeposition method

The preparation of POM EC thin films by electrodeposition also needs a stabilizer or a substrate to load the POMs; however, it is a relatively new method. This method was reported in 2013 and has obvious advantages over the LBL method.¹⁰⁸ The electrodeposition of POM films can be applied to Type 2–3 ECDs using different counter electrode materials on TCE. To meet the requirement of high transmittance at the bleached state, transparent porous TiO₂ layers have often been employed as a substrate to host POMs. In order to establish stable interactions between TiO₂ and POMs, TiO₂ should be functionalized with –OH or –COOH groups, which form chemical interactions with the terminal oxygen of POMs. Ti–OH groups can form hydrogen bonding with the terminal oxygen of POMs. Meanwhile, in the acidic condition, the Ti–OH groups are converted to a protonated form (Ti–OH₂⁺), yielding acid–base interactions with POM anions. Importantly, POMs show good chemical stability under acidic conditions except Nb- or Ta-based POMs.^120–122 Through the electrochemical technique, POMs can be deposited easily on the TiO₂ layer in acidic conditions.^123,124

The large surface area of TiO₂ is beneficial for loading large amounts of POMs as compared to that for the LBL method, reducing the deposition time. Furthermore, the porous structure is advantageous for ion diffusion, which can result in small t_c and t_b values for the EC reaction. The high transparency of TiO₂ is helpful for obtaining high ΔT, which makes the POM–TiO₂ film an ideal candidate for a POM-based EC device. A duration of less than 15 minutes was required to make a film with high performance by using the electrodeposition method.¹⁰⁸ For these reasons, POM-based EC films having various structures were prepared using this method.

4. Effect of the atomic composition and structure of POM on EC properties

Several basic structures of POMs are shown in Fig. 1. And they can be chemically manipulated to generate lacunary units by removing {MO₆} octahedra. For the same molecular structure, POM derivatives with different atomic compositions can be generated by substituting {MO₆} by other metal–oxygen octahedra. In the following section, the effect of the structure and atomic composition of POMs on the EC performance will be reviewed.

4.1 The EC properties of Dawson and lacunary Dawson-type POMs

The lacunary Dawson structures are illustrated in Fig. 1j–l. The EC properties of POM films are affected by the number of the substituted metals or the degree of lacunary structures. Under the same V_ap, the more the lacunary structures of the POM, the larger the optical contrast. Furthermore, λ_max blue-shifted with an increased degree of lacunary structures. For example, λ_max for PSS/PAH/(P₂W₁₈/PAH)₂₀ at 650 nm blue-shifted to 630 nm for PSS/PAH/(P₂W₁₇/PAH)₂₀ and to 585 nm for PSS/PAH/(P₂W₁₅/PAH)₂₀.¹²⁵ The ΔOD values of the EC films increased in the order PSS/PAH/(P₂W₁₈/PAH)₂₀ < PSS/PAH/(P₂W₁₇/PAH)₂₀ < PSS/PAH/(P₂W₁₅/PAH)₂₀ under the same applied voltage range from 0 V to −0.8 V. The ECD with PSS/PAH/(P₂W₁₅/PAH)₂₀ showed ΔA of 0.25 (Fig. 5a).

Fig. 5 (a) Solid lines: from top to bottom indicate the absorbance of ① PSS/PAH/(P₂W₁₈/PAH)₂₀ at λ_max = 650 nm, ② PSS/PAH/(P₂W₁₇/PAH)₂₀ at λ_max = 630 nm, ③ PSS/PAH/(P₂W₁₅/PAH)₂₀ at λ_max = 585 nm during double-potential steps (0 V to −0.8 V). Dotted line: the absorbance of PSS/PAH/(P₂W₁₅/PAH)₂₀ during double-potential steps from 296 to 300 cycles;¹²⁵ Reproduced from ref. 125, Copyright 2005, with permission from Elsevier. (b) P₂W₁₇/[Cu^II(phen)₂]₃₀ modified ITO during subsequent double-potential steps (−0.9V to 0.5V) at 650 nm;¹²⁸ Reproduced from ref. 128, Copyright 2005, with permission from American Chemical Society. (c) UV-vis spectra of the [P₂W₁₇/PAH/P₂W₁₇/NR]₃₅ film in the fully oxidized state (deep pink) and the fully reduced state at −0.9 V (dark purple-blue).¹²⁹ Reproduced from ref. 129, Copyright 2009, with permission from Elsevier.

According to DFT calculations,¹²⁶ all six isomeric P₂W₁₈ species have the same electronic structure constituted in the fully oxidized state, showing two well-separated sets of energy levels with a distinct energy gap (2.22–2.48 eV) between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO). The low-lying orbitals delocalize over the bridging O atoms, while the high-lying orbitals feature weak π-antibonding interactions between the metal d_xy and bridging p orbitals of O atoms.

Interestingly, the calculation for the Dawson-type POMs indicated that the polar site {WO₆} and equatorial site {WO₆} do not contribute equally to the LUMO and LUMO+1 of POMs.¹²⁶ The LUMO of these POMs is mainly centered on the equatorial site {WO₆} rather than the polar site {WO₆}. As a result, reduction occurs on the equatorial site {WO₆}, which has been confirmed experimentally.¹²⁷ This means that the reduction of POMs can be easier and better EC performances can be obtained if the equatorial site {WO₆} is more exposed. According to this principle, P₂W₁₂ can feature a high EC performance. Nonetheless, P₂W₁₂ is not stable in aqueous solutions and tends to change into P₂W₁₇.⁵⁸ P₅W₃₀ and P₈W₄₈, which are derived from P₂W₁₂, exhibit good EC properties, as will be reviewed in the following section. Since the equatorial sites {WO₆} of the Dawson-type structures and their derivatives can be exposed more, generally speaking, the lacunary and substituted Dawson structures show enhanced EC performances.⁵⁸

Several large cations, such as PSS and PAH, were employed to prepare a Dawson-type-based EC film using the LBL method. Wang et al. prepared PSS/Cu(phen)₂/[P₂W₁₇/Cu(phen)₂]₃₀ and PSS/Fe(phen)₃/[P₂W₁₇/Fe(phen)₃]₃₀ to show multicolor EC performances from light green, yellow, green to green blue (Fig. 5). PSS/Fe(phen)₃/[P₂W₁₇/Fe(phen)₃]₃₀ displayed a color change from orange red to dark violet.¹²⁸

Similarly, [P₂W₁₇/PAH/P₂W₁₇/NR]₃₅ was prepared with the help of neutral red (NR). The multilayered film showed a multicolor change from deep pink to dark bluish-purple under the potential range from 0.2 V to −1.0 V (Fig. 5c ΔA). The coloring and bleaching times were 9.3 s and 9.6 s, respectively. It showed ΔA of 0.31 unit at 650 nm under the potential steps from 0.5 V to −0.9 V.¹²⁹ In the presence of poly(vinyl alcohol) (PVA), the [PVA–P₂W₁₈/PAH]₅₀ film showed both photochromic and EC properties.¹³⁰

The P₈W₄₈ structure is a complicated lacunary Dawson-type structure. The P₈W₄₈ structure can be considered as a polymer of four P₂W₁₂ hex-vacant Dawson structures (Fig. 1l and s). Crown-type POMs (K₂₈Li₅H₇P₈W₄₈O₁₈₄·92H₂O, P₈W₄₈) and ultrathin W₁₈O₄₉ nanowires (NWs) are prepared on ITO substrates by the LBL method (Fig. 6). The prepared [PSS(PEI/PSS)₃(PEI/W₁₈O₄₉)₃₀/(PEI/P₈W₄₈)₂₀] film switched progressively between three optical states by applying different voltages. Under a positive bias, it showed transparent state (1.0 V to 0 V versus Ag/AgCl), while under a negative electrochemical bias, W₁₈O₄₉ NWs could selectively block NIR light through a plasmonic EC effect (−0.2 V to −0.6 V) (Fig. 6a, e and g). From −0.6 V to −1.0 V, the film of P₈W₄₈ absorbed a broad band of visible light (Fig. 6b, f and g).¹³¹ A smart window was fabricated using this composite film (Fig. 6d); however, an aqueous electrolyte was employed, which would be a disadvantage for long-term use. The EC properties for some examples of Dawson and lacunary Dawson-type POMs are summarized in Table 1.

Fig. 6 Transmission spectra of (a) [PSS(PEI/PSS)₃(PEI/W₁₈O₄₉)₃₀], (b) [PSS(PEI/PSS)₃(PEI/P₈W₄₈)₂₀], and (c) [PSS(PEI/PSS)₃(PEI/W₁₈O₄₉)₃₀/(PEI/P₈W₄₈)₂₀] films at different potentials. The V_ap values are 0.4 V, 0 V, −0.2 V, −0.6 V, −0.8 V and −1.0 V (from up to down) (d) schematic of an ECD; (e) the illustration of how the device realizes NIR blocking, (f) the illustration of how the device realizes dual-band blocking. (g) Power density change of the transmitted light through [PSS(PEI/PSS)₃(PEI/W₁₈O₄₉)₃₀/(PEI/P₈W₄₈)₂₀] smart window by varying V_ap. The device is at open circuit, −2.5 V for NIR-blocking state, −2.8 V for broadband-blocking state, and 2.0 V for transparent state.¹³¹ Reproduced from ref. 131, Copyright 2018, with permission from American Chemical Society.

Table 1 The EC performances of POMs with different atomic compositions and structures^a

No.	EC film	Fabrication method	Electrolyte	λ max, nm	V ap,^b V	ΔT,^c % (ΔA^d)	t b/tc^e, s	Cycle stability	Ref.
a Using Pt and Ag/AgCl as the counter and reference electrodes, respectively. b Applied voltage. c Transmittance (%). d Absorbance change. e Response time. f Sodium acetate.
1^a	PSS/PAH/(P2W15/PAH)20	LBL	HOAc–NaOAc^f buffer solution (pH 4.0)	585	−0.8/0	(0.25)	5.5/6.0	300	125
2^a	PSS/Cu(phen)2/[(P2W17/Cu(phen)2)]30	LBL	0.2 M HOAc–NaOAc^f buffer at pH = 3.5	650	−0.9/0.5	(0.24)	6.6/6.5	200	128
3^a	PSS/Fe(phen)3/[(P2W17/Fe(phen)3)]30	LBL	0.2 M HOAc–NaOAc^f buffer at pH = 3.5	650	−0.9/0.5	(0.22)	7.4/7.0	200
4^a	PSS(PEI/PSS)3(PEI/W18O49)30/(PEI/P8W48)20	LBL	0.10 M LiClO4 aqueous solution (pH = 5.0)	500	−1.0/0.4	35	26/86		131
				1060			52/86

---

4.2. Transition- and lanthanide-metal–substituted Dawson-type POMs

4.2.1 Vanadium-substituted Dawson-type POMs. In 2015, vanadium-substituted Dawson-type POM (Fig. 1m–o) EC films were prepared on porous TiO₂ using the electrodeposition method.¹³² The EC colors of the substituted POMs were affected by the degree of substitution. The tri-V-substituted P₂W₁₈ films showed a multicolor switching phenomenon from transparent to blue and bluish-purple states. Moreover, the P₂W₁₅V₃-based EC film featured the best EC performance among the three kinds of V-substituted P₂W₁₈ films. For the TiO₂(NP)–P₂W₁₇V EC film, ΔT was 85.1% under the potential step from −1.7 V to 1.5 V at 565 nm; the t_c and t_b values were 5.6 s and 4.1 s, respectively, and η_ce was 67.7 cm² C⁻¹.

For the TiO₂(NP)–P₂W₁₅V₃ EC film, ΔT at 555 nm under the V_ap step from −1.8 V to 1.3 V was 91.8% and η_ce was 176.8 cm² C⁻¹, indicating the best EC properties among those of these derivatives.¹³² The TiO₂(NP)–P₂W₁₅V₃ EC film showed color changes from transparent to blue and bluish-purple with the increase in negative potentials. However, for TiO₂(NP)–P₂W₁₇V and TiO₂(NP)–P₂W₁₆V₂, no multi-color change was detected; they only showed color changes from transparent to blue with the increase in negative potentials. This multi-color EC switching (Fig. 7) can be attributed to the electron transfer from the polar site to the equatorial site. Thus, the electrons can be first attracted by the more electronegative V⁵⁺ on the polar site of the Dawson anion and then can be transferred to the W⁶⁺ anions on the equatorial site, which are easier to be reduced.

Fig. 7 (a) Visible spectra of the α-P₂W₁₅V₃-based EC film recorded under different potentials ranging from −0.5 V to −1.8 V. (b) Visible spectra of the α-P₂W₁₇V-based EC film under different potentials ranging from −0.5 V to −1.7 V;¹³² (a and b) Reproduced from ref. 132, Copyright 2015, with permission from the Royal Society of Chemistry. (c) In situ ΔA of [P₂W₁₇Ni/PAH]₂₀ during different potentials from −0.4 V to −0.9 V.¹³³ Reproduced from ref. 133, Copyright 2005, with permission from The Electrochemical Society. (d) Absorption spectra of self-assembled ITO/(PDDA/As₂W₁₅Mg₃)₄ films at open circuit (i), −0.5 (ii), −0.6 (iii), −0.7 (iv), −0.8 (v), −0.9 (vi) and −1.0 V (vii) V_ap, inset: the response time test under open circuit to −1.0 V at 650 nm.¹³⁴ Reproduced from ref. 134, Copyright 2008, with permission from Elsevier; (e) absorbance at 585 nm of the {[Ce(P₂W₁₇)₂]/PAH}₂₀ film-modified ITO during subsequent double-potential steps (−0.9 V to 0 V)¹³⁵ (f) {[Ce(P₂W₁₇)₂]/PAH}₁₀₀ by potential turning on at −0.9 V and turning off.¹³⁵ (e and f) Reproduced from ref. 135, Copyright 2004, with permission from the Royal Society of Chemistry.

4.2.2 Other metal–substituted Dawson-type POMs. P₂W₁₇Ni exhibits a mono-substituted Dawson structure (Fig. 1m). A [P₂W₁₇Ni/PAH]₂₀ film was prepared, and its EC properties were examined at λ_max = 620 nm under the potential steps from 0 V to −0.9 V. The absorbance change was determined as 0.13 from this ECD. The response times t_c and t_b were 5.6 s and 6.4 s, respectively (Fig. 7c).¹³³ The tri-Mg-substituted Dawson structure (Fig. 1o) K₁₈[As₂W₁₅Mg₃O₆₂] was prepared by Dong's group and its EC properties were examined under the V_ap steps from 0 V to −1.0 V, showing very small ΔA (∼0.07) at λ_max = 650 nm and slow response times, i.e., t_c = 17 s and t_b = 30 s (Fig. 7d).¹³⁴ Its EC performance was inferior to that of the V-substituted Dawson-type structure possibly because the substituted metals can affect the electrochemical properties of the equatorial {WO₆} sites, which is critical for EC properties. Moreover, the EC film of As₂W₁₅Mg₃ was prepared as a dense film, which can be disadvantageous to the cation penetration and transport, resulting in low optical contrast and long response time.

In 2004, Xu et al. prepared {[Ce(P₂W₁₇)₂]}/PAH₂₀ by the LBL self-assembly method. The response times of this material were t_c = 4.0 s and t_b = 5.5 s, respectively, and ΔA at 585 nm was 0.18 under V_ap from −0.9 V to 0 V (Fig. 7e).¹³⁵ The basic structures of Ce(P₂W₁₇)₂ and P₂W₁₇Ni all consist of P₂W₁₇ blocks, while the substituted metal and substituted mode are different from each other. Compared with the result for the mono Ni-substituted Dawson (P₂W₁₇Ni) film, they show enhanced ΔA by 0.05 unit and a faster response time.

The Ln-substituted Dawson structure (Fig. 1p) also showed EC properties. In particular, the Ln-substituted series showed superior EC properties to the mono-transition metal–substituted Dawson structures. This can be observed because the Ln-substituted structures have 24 equatorial {WO₆} units, while mono-transition metal-substituted ones have 12 equatorial {WO₆} units. As described above, the more the equatorial {WO₆} units, the higher the EC activities. Therefore, {[Ce(P₂W₁₇)₂]/PAH}₂₀ showed enhanced EC performances with ΔA of 0.75 at 585 nm under the potential steps from −0.9 V to 0 V. However, the response time became very long (t_c = 108 s and t_b = 350 s, respectively) (Fig. 7f).¹³⁵

Due to their diversity in structure and atomic composition, the Dawson group of POMs can be a good EC model for understanding the structure–property relationship and for high-performance ECDs in the future. Some of the substituted POMs exhibit their intrinsic colors due to the presence of substituted metal ions. For example, the V-substituted and the Ni-substituted Dawson-type structures present orange and light green colors, respectively. Furthermore, the Ln-substituted Dawson-type structures show photoluminescence,^136,137 which can be interesting for multi-functional ECDs. The EC properties of the transition and lanthanide metal–substituted Dawson-type POMs are summarized in Table 2.

Table 2 The EC properties of the transition and lanthanide metal–substituted Dawson-type POMs

No.	EC film	Fabrication method (thickness)	Electrolyte	λ max, nm (at Vap^b)	V ap,^b V	ΔT,^c % (ΔA^d)	t b/tc,^e s (ηce,^f cm^2 C^−1)	Cycle stability	Ref.
a Using Pt and SCE as the counter and reference electrodes, respectively. b Applied voltage. c Transmittance (%). d Absorbance change. e Response time. f Coloration efficiency. g Using Pt and Ag/AgCl as the counter and reference electrodes, respectively. h Sodium acetate.
5^a	TiO2(NP)–P2W15V3	Electrodeposition (6 μm)	0.1 M HCl aqueous solution	555 (−1.4 to −1.8)	−1.8/1.3	91.8	3.5/3.9 (177)	1000	132
				577 (−1.3)
				586 (−0.8 to −1.2)
6^a	TiO2(NP)–P2W16V2	Electrodeposition (6 μm)	0.1 M HCl aqueous solution	590	−1.6/1.3	48.3	11.1/4.5 (47.8)	1000
7^a	TiO2(NP)–P2W17V	Electrodeposition (6 μm)	0.1 M HCl aqueous solution	565	−1.7/1.5	85.1	5.6/4.1 (67.7)	1000
8^g	{[Ce(P2W17)2]/PAH}20	LBL	0.1 M HOAc–NaOAc^h buffer solution (pH ∼ 4.0)	585	−0.9/0	(0.18)	4.0/5.5	200	135
9^g	{[Ce(P2W17)2]/PAH}100	LBL				(0.75)	108/350	—

---

4.3. Preyssler-type POMs

Preyssler-type POMs showed a high EC performance, which could be attributed to the high ratio of equatorial {WO₆} units in the structure (Fig. 1r). These Preyssler-type structures have more negative charges, which result in the enhancement of the POM coverage. Preyssler-type POMs have more negative charges than the Keggin- and Dawson-type POMs. Thus, Preyssler-type POMs are more effective species to make LBL films.

In 2002, Liu et al. first identified the EC behavior of the Preyssler structure (Eu-P₅W₃₀:EuP₅W₃₀O₁₁₀¹²⁻). The film of (PSS/PAH/Eu-P₅W₃₀/PAH)₂₀ showed ΔA of 0.12 at 700 nm under the potential step from −0.4 V to −1.8 V. When the number of double layers reached 100, ΔA was enhanced to 0.7; however, the response time became very long up to 7 min (Fig. 8a).¹³⁸ In 2006, ΔA of the film of PSS(P4VP/Na-P₅W₃₀)₄₀ (Na-P₅W₃₀:NaP₅W₃₀O₁₁₀¹⁴⁻) with poly(4-vinylpyridine) (P4VP) was reported as 0.28 under V_ap of −2.1 V/0 V.¹³⁹

Fig. 8 (a) ΔA at 700 nm of the (PSS/PAH/Eu-POM/PAH)₂₀-coated ITO electrode during subsequent double-potential steps between −0.4 V and −1.8 V, showing the ΔA switching at the 1st (solid line) and 500th (dashed line) cycle; traces are offset for clarity. The response times of t_c and t_b were 4.2 and 4.4 s, respectively.¹³⁸ Reproduced from ref. 138, Copyright 2002, with permission from John Wiley & Sons, Inc. (b) Illustration of the ECD structure of the X₅W₃₀ (X = P, S)-based smart windows. ① Photographic images of the bleached ECD, ② the P₅W₃₀-based ECD under −1.5 V, and ③ the S₅W₃₀-based ECD under −2.0 V. (c) At 650 nm, ΔT for the P₅W₃₀-based ECD during the subsequent double-potential steps between 0.5 V and −1.5 V.¹⁰⁸ (b and c) Reproduced from ref. 108, Copyright 2013, with permission from the Royal Society of Chemistry.

The films of TiO₂(NP)–Na-P₅W₃₀ and TiO₂(NP)–Na-S₅W₃₀ (Na-S₅W₃₀:NaS₅W₃₀O₁₁₀⁹⁻) prepared by electrodeposition showed good EC performances in a Type 2a structure (Fig. 8). The response time of the TiO₂(NP)–Na-P₅W₃₀ film was less than 2 s for both t_c and t_b values along with ΔT of 50% under the V_ap step from −1.5 to 0.5 V at 650 nm (Fig. 8c). The TiO₂(NP)–Na-S₅W₃₀ film showed small ΔT of 20% under V_ap from −2.0 to 0.5 V at 550 nm. Obviously, the S₅W₃₀ derivative showed worse EC properties than the P₅W₃₀ derivative. This could be attributed to the S atom having higher electronegativity than the W atom, which makes {WO₆} harder to be reduced in the S₅W₃₀ derivative.¹⁰⁸ This suggests that the heteroatoms also have a great influence on the EC properties. Moreover, the porous morphology of WE contributed significantly to the enhancement in the performance. The EC properties of the Preyssler-type and isopolyoxometalates are compared in Table 3.

Table 3 The EC properties of the Preyssler-type and isopolyoxometalates

No.	EC film	Fabrication method (thickness)	Electrolyte	λ max, nm	V ap,^b V	ΔT,^c % (ΔA^d)	t b/tc,^e s (ηce,^f cm^2 C^−1)	Cycle stability	Ref.
a Using Pt and SCE as the counter and reference electrodes, respectively. b Applied voltage. c Transmittance (%). d Absorbance change. e Response time. f Coloration efficiency. g Using Pt and Ag/AgCl as the counter and reference electrodes, respectively. h Using FTO as the CE in 2-electrode device. i Lithium iodide. j Propylene carbonate.
10^a	(PSS/PAH/Eu-P5W30/PAH)20	LBL	pH 3.0 buffer solutions	700	−0.4/−1.8	(0.09)	≤420	500	138
11^a	(PSS/PAH/Eu-P5W30/PAH)100	LBL				(0.7)	1800
12^g	(PDDA/W10)n (n = 3, 6, 10)	LBL	0.1 M aqueous NaCl at pH = 2.5	650	−0.8/0.4	(0.025)	5/65(3)	60	140
							15/70(6)
							20/80(10)
13^a	TiO2(NP)–W10	Electrodeposition (4 μm)	0.1 M H2SO4 aqueous solution	550	−0.8/0.5	49.5	7.2/4.9	100	141
14^h	TiO2(NP)–Na-P5W30	Electrodeposition (4 μm)	0.1 M LiI^i/PC^j	650	−1.5/0.5	50	1/1 (711)	1000	108
15^h	TiO2(NP)–Na-S5W30	Electrodeposition (4 μm)	0.1 M LiI^i/PC^j	550	−2/1	23		1000

---

4.4. Isopolyoxometalate

The isopolyoxometalate (iso-POM) species is also a class of POMs, and it shows good EC properties. The most representative and simplest iso-POM is the W₁₀ (W₁₀O₃₂⁴⁻) structure (Fig. 9a). (PDDA/W₁₀)₆ was prepared in 1998. Under V_ap from −0.8 V to open circuit, it shows small ΔA of 0.02 at 650 nm.¹⁴⁰ The W₁₀-based EC film was also prepared through electrodeposition on a porous TiO₂ substrate, and it showed enhanced performance compared to the film prepared by the LBL method. The response times t_c and t_b were 7.2 s and 4.9 s, respectively, with ΔT of 47.1% at 550 nm under V_ap from 0.5 V to −0.8 V (Fig. 9b).¹⁴¹

Fig. 9 (a) The structure of W₁₀; (b) the transmittance spectra of the FTO, TiO₂ substrate and the EC film under different V_ap.¹⁴¹ Reproduced from ref. 141, Copyright 2017, with permission from Elsevier.

5. High-performance EC film based on TiO₂ nanoparticle-POMs

TiO₂ nanoparticles (NP) are widely used semiconductors in various areas and they have been demonstrated to be good stabilizers for POMs. The TiO₂(NP)–POM composite film has been prepared through various processes such as sol–gel and impregnation methods.^142–149 To prepare a transparent composite film for an EC application, a two-step process is often employed: (1) preparation of the TiO₂ nanostructure dispersion or film and (2) the combination of POMs with TiO₂. The hydrogen bond between the terminal oxygen of POMs and the surface hydroxyl groups (Ti–OH) of TiO₂ is one reason for the combination of POMs and TiO₂. Another reason is that under acidic conditions, the protonated TiO₂ (Ti–OH₂⁺) easily combines with POMs through an acid–base reaction, which can be realized by LBL and electrodeposition methods.¹⁵⁰

Based on the morphology of TiO₂ and the film preparation process, TiO₂(NP)–POM composite films with different morphologies were prepared (Fig. 10). First, TiO₂(NP) with a diameter of 17.3 nm was used to prepare multilayer POM, yielding a flat condensed film (Fig. 10a). The [P₂W₁₈/PAH/P₂W₁₈/TiO₂]₁₀ multilayer film showed coloring and bleaching times of 15.4 s and 13.4 s, respectively, and ΔA at 650 nm was 0.12 units under V_ap from 0.5 V to −0.9 V (Fig. 10b). The [P₂W₁₈/PAH/P₂W₁₈/TiO₂]₁₀ multilayer film achieved longer stability than the [P₂W₁₈/PAH]₂₀ film after 1000 cycles of double potential steps (Fig. 10c). However, TiO₂ did not significantly affect other properties such as response time and ΔT.⁷⁶

Fig. 10 (a) SEM image of a [P₂W₁₈/PAH/P₂W₁₈/TiO₂]₁₀ multilayer film on a silicon wafer; (b) visible spectra of [P₂W₁₈/PAH/P₂W₁₈/TiO₂]₁₀-modified ITO glass electrode at different potentials. (c) The absorption changes for the double-potential step at a = 1, b = 500 and c = 1000 double-potential steps;⁷⁶ (a–c) Reproduced from ref. 76, Copyright 2009, with permission from Elsevier. (d) The SEM images of the TiO₂(NP)–P₂W₁₇ film on FTO; (e) the illustration of the structure of the POM-based smart windows; (f) the photos of the POM-based smart windows at bleaching and coloration states. (g) Visible spectra of the P₂W₁₇-based EC smart window under different potentials ranging from −0.8 V to −1.8 V. Inset: Plots of absorbance (at 620 nm) versus the V_ap;⁴⁹ (d–g) Reproduced from ref. 49, Copyright 2013, with permission from Elsevier. (h) SEM images of TiO₂–[P₂W₁₈/PEI]₄₀ films, inset: cross-sectional SEM image of TiO₂-[P₂W₁₈/PEI]₄₀ film; (i) schematic of formation of a TiO₂(NW)–P₂W₁₈ film; (j) in situ transmittance curves for TiO₂(NW)–[P₂W₁₈/PEI]₄₀, [P₂W₁₈/PEI]₄₀ and TiO₂(NW) films at 650 nm under the potential step from 1.5 V to −1.5 V.¹⁵¹ (h–j) Reproduced from ref. 151, Copyright 2017, with permission from Elsevier.

Second, Dawson-type POM films (TiO₂(NP)–P₂W₁₈, TiO₂(NP)–P₂W₁₇) were prepared through electrodeposition on a porous TiO₂ substrate with 25 nm particle diameter. The film featured porous morphology, which would be beneficial for the fast diffusion of Li⁺. The composite films were assembled as Type 2b ECD using 0.1 M LiI propylene carbonate as the electrolyte and FTO as CE; the configuration is shown in Fig. 10e. For the P₂W₁₈–TiO₂ film, ΔT was 48.7% under V_ap of −1.9 V at 646 nm; the t_c and t_b values were 0.97 s and 1.98 s, respectively, and η_ce was 176.8 cm² C⁻¹. This was nearly 10 times faster than the response time reported for [P₂W₁₈/PAH/P₂W₁₈/TiO₂]₁₀, which was possibly due to the optimized morphology of the TiO₂(NP)–P₂W₁₈ film.⁴⁹

Interestingly, the TiO₂(NP)–P₂W₁₇ film showed high ΔT of 85.1% at the V_ap step from −1.7 V to 1.0 V at 620 nm (Fig. 10g). The response times t_c and t_b of the TiO₂(NP)–P₂W₁₇ film were 0.90 s and 2.81 s, respectively, along with high η_ce of 205.3 cm² C⁻¹. Furthermore, the TiO₂(NP)–P₂W₁₇ film showed a fast response time and high EC properties at near IR region (λ_max = 1050 nm) under the V_ap steps from −1.7 to 1.0 V with a contrast of 62.3%. These results also confirmed the principle discussed on the lacunary Dawson structures, which exhibit better EC performances than the saturated Dawson structures.

Third, in 2017, Liu et al. modified the LBL process for preparing the POM film, which was carried out on the surface of the TiO₂ nanowire (NW) with the length of ∼600 nm and the diameter of ∼50 nm. TiO₂(NWs) were covered by POMs via the LBL process; however, the pores between TiO₂(NWs) were maintained to provide high surface areas and porosity. Under the V_ap step from 1.5 V to −1.5 V, TiO₂(NW)–[P₂W₁₈/PEI]₄₀ showed ΔT of 45.1% at 650 nm with the response times of t_c = 1.9 s and t_b = 6.7 s (Fig. 10j).¹⁵¹

The P₂W₁₈ films with different morphologies were examined to reveal their morphology-dependent EC performance. The EC films having nanoparticle and nanowire structures exhibited nearly the same optical density around 50%, which was higher than that of the LBL film; this was mainly due to the higher content of P₂W₁₈ in NP and NW films than that in LBL films. The response time for the NP and NW films was significantly faster than that for the LBL film. The NP particle films featured larger pore sizes than the NW film, which would be beneficial for fast Li⁺ intercalation/deintercalation, resulting in a fast response time. The effects of the morphology on the EC response time were more obviously correlated between the porous films and the dense LBL film. Table 4 summarizes the EC properties of the TiO₂–POM composite film.

Table 4 EC properties of the TiO₂–POM composite film

No.	EC film	Fabrication method (thickness)	Electrolyte	λ max, nm	V ap,^b V	ΔT,^c % (ΔA^d)	t b/tc,^e s (ηce,^f cm^2 C^−1)	Cycle stability	Ref.
a Using Pt and SCE as the counter and reference electrodes, respectively. b Applied voltage. c Transmittance (%). d Absorbance change. e Response time. f Coloration efficiency. g Using Pt and Ag/AgCl as the counter and reference electrodes, respectively. h Using FTO as the CE in 2-electrode device. i Lithium iodide. j Propylene carbonate. k Sodium acetate.
5^a	TiO2(NP)–P2W15V3	Electrodeposition (6 μm)	0.1 M HCl aqueous solution	555 (−1.4 to −1.8)	−1.8/1.3	91.8	3.5/3.9 (176.8)	1000	132
				577 (−1.3)
				586 (−0.8 to −1.2)
6^a	TiO2(NP)–P2W16V2	Electrodeposition (6 μm)	0.1 M HCl aqueous solution	590	−1.6/1.3	48.3	11.1/4.5 (47.8)	1000
7^a	TiO2(NP)–P2W17V	Electrodeposition (6 μm)	0.1 M HCl aqueous solution	565	−1.7/1.5	85.1	5.6/4.1 (67.7)	1000
13^g	TiO2(NP)–W10	Electrodeposition (4 μm)	0.1 M H2SO4 aqueous solution	550	−0.8/0.5	49.5	7.2/4.9	100
14^h	TiO2(NP)–Na-P5W30	Electrodeposition (4 μm)	0.1 M LiI^i/PC^j	650	−1.5/0.5	50	1/1 (71.1)	1000	108
15^h	TiO2(NP)–Na-S5W30	Electrodeposition (4 μm)	0.1 M LiI^i/PC^j	550	−2/1	23		1000
16^a	[P2W18/PAH/P2W18/TiO2]10	LBL (5 μm)	pH 2.0 HCl solution	650	−0.9/0.5	(0.12)	15.4/13.4	1000	76
17^g	TiO2-(NW)–[P2W18/PEI]40	LBL (600 nm)	0.2 M, HOAc–NaAc^k pH = 3.4	600	−1.5/1.5	45.1	1.9/6.7 (69)		151
18^a	TiO2(NP)–PW11V	Electrodeposition (∼4 μm)	0.1 M HCl aqueous solution	650	−1.5/1.5	75.8	15.6/8.7	1000	152
	TiO2(NP)–PW10V2					74.5	17.2/6.4
	TiO2(NP)–PW9V3					79.7	7.1/2.4
19^a	TiO2(NP)–PW11V	Electrodeposition (∼4 μm)	0.1 M LiClO4 PC^j solution	660	−2.0/1.8	44.7	19.5/4.3	1000
	TiO2(NP)–PW10V2					66.3	6.7/5.2
	TiO2(NP)–PW9V3					77.6	13.6/6.2

---

6. POM composite films with other metal oxides and carbon materials

6.1 POM composite films with other metal oxides

Besides TiO₂, other metal oxides or semiconductors can also be used as stabilizers or linkers with POMs to prepare EC films. CdS nanoparticles (size: 11.6 nm) are employed to enhance the performance of the P₂W₁₇V-based EC films. Ma et al. prepared the {PEI/[P₂W₁₇V/PEI–CdS]₃₀/P₂W₁₇V} film to show ΔT of 59.8% optical contrast at 573 nm along with t_c/t_b as 5.6/2.4 s under the V_ap step from −1.1 V to 0 V. The CdS nanoparticles contributed to a large surface area, good conductivity, fast IETs with P₂W₁₇V and fast charge compensation by ions in the contacting liquid, which remarkably enhanced the EC performance of the composite film in comparison to that of the CdS-free film.¹⁵³ In 2016, the composite film of WO₃ sol (negatively charged) with P₂W₁₈ and PEI was prepared by the LBL method. WO₃ and P₂W₁₈ synchronously contributed to the improvement in the EC properties. The optical contrast of the [PEI/P₂W₁₈/PEI/WO₃]₂₀ film at 650 nm under the polarization of −1.5 V/1.5 V was 48.4%, which was higher than those of [PEI/P₂W₁₈]₄₀ (23.5%) and [PEI/WO₃]₄₀ (14.0%) films. The synergetic effect was attributed to the large surface area, which facilitated ion intercalation and de-intercalation.¹⁵⁴

Among the studies on POMs, the studies on Keggin-type POMs are far lesser than those on Dawson-type POMs. In contrast to the LUMO of the Dawson structures, the LUMO of Keggin-type structures is located on the polar sites according to DFT calculations.¹⁵⁵ Therefore, a lacunary Keggin structure would have less EC-active units, which will not be beneficial for EC performance. However, in the saturated and substituted Keggin structures, the substituted atoms significantly affect the LUMO and lead to an enhanced EC performance.^156,157 Therefore, the saturated and substituted Keggin structures are the best choices for EC materials among the Keggin structures.

Recently, SnO₂ nanoparticles were used as the substrate to absorb PMo₁₂ through the LBL process. Due to the large surface area, nano-SnO₂/PDDA/PMo₁₂ film was only deposited once. The conductivity of the film became higher with the help of SnO₂. The response time was enhanced considerably; however, the optical contrast was still low (ΔT = 11%).¹⁵⁸

The substituted Keggin-type structures are shown in Fig. 1e–g. In 2009, the SiW₉V₃ and Bi₂O₃ composite film was prepared using the LBL method. PEI-coated Bi₂O₃ nanoparticles were used as the polycations. The EC response times of the PEI/[SiW₉V₃/PEI-Bi₂O₃]₁₅/SiW₉V₃ film were reported to be relatively fast: 7.0 s and 6.3 s for t_c and t_b, respectively. However, ΔA was low (0.03) under the potential steps from −1.0 V to 0 V at 678 nm (Fig. 11a). Nonetheless, the Bi₂O₃ nanoparticles incorporated in the films have the ability to shorten the response time and enhance the efficiency of photochromism.¹⁵⁹

Fig. 11 (a) Absorbance at 678 nm of the {PEI/[SiW₉V₃/PEI-Bi₂O₃]₁₅/SiW₉V₃} film-modified ITO during subsequent double-potential steps (−1.0 V to 0 V);¹⁵⁹ Reproduced from ref. 159, Copyright 2009, with permission from American Chemical Society. (b) Visible spectra of the α-PW₁₁V-based film in H⁺-based aqueous electrolyte under different V_ap and (c) Li⁺-based non-aqueous electrolyte under different V_ap.¹⁵² (b and c) Reproduced from ref. 152, Copyright 2016, with permission from the Royal Society of Chemistry.

V-Substituted Keggin-type POMs PW_12−nV_n (n = 1, 2, 3) were prepared using the electrodeposition method. These films exhibited a good EC performance in both acid-aqueous and non-aqueous Li⁺/PC electrolytes (Fig. 11b and c).¹⁵² Most of the research on POM EC films has been carried out in acidic aqueous electrolytes. However, non-aqueous electrolytes would be beneficial for the stabilization of EC materials and devices.

6.2. POMs with carbon material and polymers

The combination of POMs with carbon materials and polymers is another attractive way to make POM-based EC films. There are several ways to realize non-covalent combination between carbon materials and POMs. First, using electrostatic interactions, many POM–carbon nanotubes (CNTs) and POM–graphene hybrids have been prepared (Fig. 12a).^160–165 Thin films are prepared using the LBL method, which is also based on the electrostatic interactions (Fig. 12b).^166–168 Second, π–π interaction with carbon materials affords a POM hybrid composite. This requires POMs functionalized with aromatic groups, which generates π–π interaction of POMs with the surface of CNTs or graphene (Fig. 12c).^169,170

Fig. 12 (a) The illustration of the POM cluster SWCNT composite.¹⁶⁵ Reproduced from ref. 165, Copyright 2016, with permission from the Royal Society of Chemistry. (b) The illustration of POM cluster combined with polycation-decorated MWCNT.¹⁶⁴ Reproduced from ref. 164, Copyright 2010, with permission from Springer Nature. (c) Schematic of POM-based hybrids to form the π–π interaction with the surface of the CNT or graphene.¹⁷⁰ Reproduced from ref. 170, Copyright 2013, with permission from John Wiley & Sons, Inc. (d) The SEM images of [P₂W₁₈/CS-CNTs]₃₀ film (up) and [P₂W₁₈/CS]₃₀ film (down) on ITO-coated glass; (e) UV-vis spectra of [P₂W₁₈/CS-CNTs]₃₀ and [P₂W₁₈/CS]₃₀ films under a constant potential of −0.9 V; (f) ΔOD versus the extracted charge density of [P₂W₁₈/CS-CNTs]₃₀ and [P₂W₁₈/CS]₃₀ films.¹⁷¹ (d–f) Reproduced from ref. 171, Copyright 2011, with permission from Elsevier.

The pyrene-based compound is a good linker between POM and CNTs. There are two ways to use a pyrene-based compound as a linker. The lacunary POM site can be covalently functionalized with a pyrene compound and then, pyrene-modified POM hybrids can be used for the non-covalent sidewall functionalization of single-walled CNTs (SWCNTs). For example, 1-pyrenemethylamine hydrochloride (PMA) was reported as a linker between CNTs and P₂W₁₈. In this sample, P₂W₁₈ structures were well-dispersed on the CNT surface to provide effective interfacial electron transfer (IET) between CNT and P₂W₁₈.¹¹⁴

Surfactants are also used as linkers between POMs and CNTs. Hong's group used cellulose sodium salt and Triton X-100 as the linkers to adsorb PMo₁₂ on SWNTs. Efficient IET occurs from nonionic Triton X-100/CNT to PMo₁₂. However, with the anionic surfactant (cellulose sodium), there is no IET between them.¹⁸ In 2011, as an alternative stabilizer for POMs, Xu's group used chitosan (CS) as a linker to form a stable P₂W₁₈/CS-CNTs composite film by the LBL method. The film with CNTs showed porous morphology and the film without CNT showed a flat dense surface. Under the potential step from −0.9 V to 0.5 V, η_ce of the [P₂W₁₈/CS-CNTs]₃₀ film at 620 nm was 91.5 cm² C⁻¹, while η_ce for [P₂W₁₈/CS]₃₀ was 74.7 cm² C⁻¹, indicating that CNTs can assist in enhancing the performance of the EC film. Nonetheless, the optical contrast was still low (20.3%) (Fig. 12d–f).¹⁷¹

The [P₂W₁₈/PAH/P₂W₁₈/PEDOT]₂₀ composite film was prepared by the combination of LBL and electropolymerization. Here, PEDOT could be regarded as a linker for P₂W₁₈. The optical contrast of the [P₂W₁₈/PAH/P₂W₁₈/PEDOT]₂₀ film was 54.7% under a potential step from −1.5 V to 1.0 V, which was higher than those of the [P₂W₁₈/PAH]₄₀ (42.3%) and electropolymerized PEDOT (20.3%) films. Both P₂W₁₈ and PEDOT synchronously contributed to the enhancement in the EC properties.¹⁷² The EC properties of the POM composite films with other metal oxides and carbon materials are summarized in Table 5.

Table 5 EC properties of the POM composite films with other metal oxides and carbon materials

No.	EC film	Fabrication method (thickness)	Electrolyte	λ max, nm	V ap,^b V	ΔT,^c % (ΔA^d)	t b/tc,^e s (ηce,^f cm^2 C^−1)	Cycle stability	Ref.
a Using Pt and SCE as the counter and reference electrodes, respectively. b Applied voltage. c Transmittance (%). d Absorbance change. e Response time. f Coloration efficiency. g Using Pt and Ag/AgCl as the counter and reference electrodes, respectively. h Phosphate buffered saline. i Propylene carbonate. j Sodium acetate.
20^g	{PEI/[P2W17V/PEI–CdS]n}/P2W17V	LBL (411 nm for 20 layers)	0.05 M NaH2PO4–Na2HPO4 buffer solutions (PBS,^h pH = 7.0)	573	−1.1/0	22.7, 45.3, 59.8 (0.11, 0.26, 0.39)	3.2/1.9 (24.33)		153
							5.6/2.1 (29.77)
							5.6/2.4 (38.29)
21^g	[PEI/P2W18/PEI/WO3]20	LBL (20 layers, ∼1 μm)	1 mol L^−1 LiClO4 in PC^i	650	−1.5/1.5	48.4	24.9/2.4 (46.1)		154
22^g	Nano-SnO2/PDDA/[PMo12O40]	LBL (2.1 μm)	0.1 M Na2SO4 (pH = 5.5)	700	−0.7/0.2	11 (0.045)	1/1 (3.5)		158
23^a	PEI/[SiW9V3/PEI–Bi2O3]15/SiW9V3	LBL (15 layers)	0.2 M HOAc–NaOAc^j (pH = 3.82) buffer solutions.	678	−1.0/0	(0.03)	7.0/6.3 (7.8)	200	159
24^g	[P2W18/PAH/P2W18/PEDOT]20	LBL, with PEDOT (6 μm)	0.1 M LiClO4 in Na2SO4–H2SO4 buffer solution (pH = 3)	600	−1.5/1.0	54.7	2.45/1.41	20 (1.0/−1.5 V); 1500 (1/−1 V)	172
25^g	[P2W18/CS-CNTs]30	LBL	HOAc–NaOAc^j buffer solution (pH = 4)	620	−0.9/0.5	20.3	4.3/3.7 (91.5)	1000	171

---

7. POM composite films for bistable and capacitive EC windows

EC devices with POMs generally consist of a POM-coated WE, an electrolyte layer, and a CE (like Type 2b), as described above. In order to improve the EC properties and add additional functionalities, CE can be coated with other EC materials (Type 3a) or with charge balancing materials (Type 3b). Furthermore, for a complete EC reaction, a full-cell structure is necessary to reach charge balancing between WE and CE. Thus, the performance of ECDs with CBL on CE would be better than that of ECDs without CBL. In Type 3a, EC reactions can be charge balanced with counter EC materials; however, the optical contrast would be worse than that in Type 3b because of the intrinsic color of the EC materials of CE in Type 3a.

A recent study has revealed that transparent TiO₂ thin films can provide charge balancing property for the EC reaction of a CP.⁶ In this study, the CP underwent cathodically coloring EC switching according to the following reactions:

(Pol*) + n(A⁻C⁺)s ⇄ (Pol⁺ⁿA_n⁻) + ne⁻ + nC⁺ (5)

Colored colorless

(CE*) + ne⁻ + nC⁺ ⇄ (CE⁻ⁿC_n⁺) (6)

In the EC potential range for CP (eqn (5)), the TiO₂ layer coated on CE did not show any color change but it behaved as an electron reservoir (eqn (6)). The electrode potential reached at WE for the EC reaction (eqn (5)) of CP in this type 3b is dependent on the thickness of the TiO₂ layer according to the modified Nernst equation given in eqn (7):Here, E_we is the potential developed on WE under a given applied potential (E_ap); is the formal potential of TiO₂; n is the number of electrons transferred from the EC film to the external circuit; f and t are the fractions of the reduced TiO₂ and the thickness of the TiO₂ film, respectively; F, R, and T are the Faraday constant, universal gas constant and absolute temperature, respectively; and η_CE and iR_ce are the potential drop caused by the ohmic resistance at CE and the overvoltage on CE, respectively. As a result, E_we was enhanced by increasing the thickness of the TiO₂ layer, which functioned as CBL. In this manner, the working voltage of ECD could be reduced and overvoltage could be determined via a dual potential measurement system. Although this system requires a thick TiO₂ layer, TiO₂ nanoparticle layers cannot be made as a thick film due to their poor electrical conductivity. Instead, it can be coated with redox-active species such as POMs.

Since POMs can reversibly accept and donate electrons without structural changes, they can act as CBL or CBL assistant. In particular, with TiO₂ as the stabilizer, it can shuttle the redox reaction for EC switching and improve the EC properties of a conjugated polymer-based ECD, thus achieving bistable ECD and ECCW, as described below.¹⁷³

The LUMO level of PW₁₂ is slightly lower than the conduction band of TiO₂.¹⁷⁴ Thus, when combined with TiO₂, the reduced TiO₂ can transfer electrons to PW₁₂, which functions as a redox shuttle. To explore the charge balancing effect, a TiO₂(NP)–PW₁₂ composite for the EC polymer devices, PW₁₂, was electrodeposited on a TiO₂-coated electrode using the reduction of W(VI) (eqn (8)) through the cyclic voltammetry (CV) method.¹⁷³ SEM and energy-dispersive spectroscopy (EDS) mapping of the cross-section of the TiO₂(NP)–PW₁₂ samples with different TiO₂ thicknesses showed that PW₁₂ was uniformly coated not only on the surface but also in the pores between the TiO₂ nanoparticles as an ultrathin film over the entire TiO₂ layer (Fig. 13).

n[PW^VI₁₂O₄₀³⁻] + nTiO₂(e⁻) + nH⁺ → n[HPW^V₁₂O₄₀³⁻–TiO₂] (8)

The CV of TiO₂(NP)–PW₁₂ clearly presents characteristic oxidation peaks at −0.41 V and −0.27 V versus Ag/AgCl in an aqueous electrolyte containing LiClO₄ (0.1 M); they were linearly correlated to the thickness of TiO₂(NP)–PW₁₂ (Fig. 13h, inset). These two peaks were attributed to the two steps of the one-electron process as follows:¹⁷⁵

PW₁₂O₄₀³⁻ + e⁻ → PW₁₂O₄₀⁴⁻ (9)

PW₁₂O₄₀⁴⁻ + e⁻ → PW₁₂O₄₀⁵⁻ (10)

Fig. 13 (a) Illustration of PW₁₂ coating on TiO₂. (b and c) Top-view SEM images of (b) 1930 nm thick TiO₂ layer (4) and (c) PW₁₂ composite (4p). (d–g) Cross-section SEM images of TiO₂(NP)–PW₁₂ layer with the TiO₂ thicknesses (t) of 633 nm (1p), 845 nm (2p), 1190 nm (3p), and 1930 nm (4p) (row i, respectively; rows ii–iv): EDS mapping of (ii) O, (iii) Ti, and (iv) W elements on the cross-section of 1p–4p. (h) CV curves of 1p–4p on FTO in an aqueous solution of LiClO₄ (0.1 M) (V versus Ag/AgCl) with a scan rate of 100 mV s⁻¹ using Pt sheet as a CE. The inset shows the correlation of the peak current density against the PWTNF thickness. (i) Scheme of ECCW with PR-Br (180 nm)-coated ITO electrode (WE); PW₁₂-coated TNF on FTO (CE) with different thicknesses, BMIM-TFSI (electrolyte) layer with a 100 μm-thick spacer.¹⁷³ Reproduced from ref. 173, Copyright 2019, with permission from John Wiley & Sons, Inc.

The redox peaks for TiO₂ appeared in a more negative range. The redox reaction for the TiO₂(NP)–PW₁₂ film-coated electrode can be described below (eqn (11)):

TiO₂ + e⁻ ⇄ TiO₂(e⁻) (11)

It is well-known that the POMs exhibit cathodically coloring electrochromism in H⁺- and Li⁺-based electrolytes via the insertion of these small ions into reduced POMs.^108,132 However, POMs did not show any color or transmittance changes when subjected to potential cycles from −1.0 V to 2.0 V (versus Ag/AgCl) in 1-butyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide (BMIM-TFSI) (Fig. 13b(ii) and (iv)) possibly due to the large BMIM cations (6.5 Å),¹⁷⁶ which cannot intercalate with POMs. It has been reported^{75,108,132,177–179} that POMs cannot show obvious color change if there is no small cation (such as H⁺ and Li⁺) intercalation. Thus, the above TiO₂(NP)–PW₁₂-coated electrode can be safely used as a non-EC charge balancing electrode without causing unwanted loss of the color contrast for CP-based ECD.

A type 3b EC window was prepared with five layers (Fig. 13i): the EC polymer poly(3,3-bis(bromomethyl)-3,4-dihydro-2H-thieno[3,4-b][1,4]dioxepine)s (PR-Br)-coated electrode (working electrode, WE); the TiO₂(NP)–PW₁₂ film-coated electrode (CE); and the electrolyte layer sandwiched between WE and CE. The 180 nm-thick PR-Br layer was used as the EC layer (eqn (5)) and BMIM-TFSI was used as the electrolyte layer.

ECD shows a large transmittance change between −0.2 V and 1.2 V (versus Ag/AgCl) (Figure 14b(iii)). Interestingly, T_b was highly dependent on the TiO₂(NP)–PW₁₂ film composition, which affected the switching voltage and the optical contrast of the ECD. Therefore, even in the low switching voltage condition, ECD with optimized TiO₂(NP)–PW₁₂ film composition showed maximum T_b of 70% or higher. Moreover, in the presence of PW₁₂, it was possible to reduce the thickness of the TiO₂ film at CE by ≈2.3 times, which could eventually decrease the thickness of ECD.

In a two-electrode EC system, V_ap is distributed between WE and CE. It is important to obtain the highest potential at WE (E_pol) with the lowest V_ap value to maximize the optical contrast with long-term stability. E_pol depends on V_ap and E_ce, which must be originating from the charge-balancing capacity of TiO₂ and PW₁₂, as presented in eqn (12) and (13).

TiO₂*_film + ne⁻ ⇄ TiO₂*(e⁻)_film (12)

[PW^VI₁₂O₄₀³⁻]_n + TiO₂(e⁻)_film ⇄ [PW^VI₁₂O₄₀³⁻–TiO₂(e⁻)]_film (13)

The fraction of charge balancing species at CE can be represented for the two species PW₁₂ and TiO₂ by the thickness of the TiO₂ layer (t) and the content of PW₁₂(C_PW12) in the TiO₂ layer, respectively. Then, the Nernst equation can be modified as eqn (14) and (15):

Here, E_pol and E_ce are the potentials on WE and CE under V_ap; and are the formal potentials of TiO₂ and PW₁₂, respectively; n₁ and n₂ are the electrons accepted by TiO₂ and PW₁₂, respectively; n₁ + n₂ is equal to the number of electrons (n) transferred from PR-Br to ITO; and f_TiO2 and f_PW12 are the fractions of TiO₂ and PW₁₂ in CE.

The energy levels of PW₁₂ are compared with those of PR-Br and TiO₂ in Fig. 14.¹⁷³ The LUMO values of TiO₂ and PW₁₂ are −4.0 and −4.73 eV, respectively. Thus, electrons from TiO₂ can be transferred efficiently to PW₁₂ and TiO₂ during the charge-balancing reaction between WE and CE. Although only a small amount of PW₁₂ (4.8–14.9 μg) is present in TiO₂, the PW₁₂ effect in ECD is significantly enhanced (Fig. 15a). Only 1.2 μm thickness of TiO₂(NP)–PW₁₂ is needed for the best performance of ECD; however, without PW₁₂, for the best performance of the ECD, a 3.4 μm TiO₂ film should be used as CBL. In other words, a small amount of PW₁₂ in TiO₂ results in the increase in the thickness and weight of TiO₂ for the charge-balancing electron transport. This indicates that the ultrathin PW₁₂ coatings function as a redox shuttle for deriving high-color-contrast ECDs.¹⁷³ Due to the optimized charge-balancing, the cyclic stability of the PW₁₂-containing ECDs notably increased (Fig. 15d).

Fig. 14 A mechanism of the charge-balancing effect by PW₁₂.¹⁷³ Reproduced from ref. 173, Copyright 2019, with permission from John Wiley & Sons, Inc.

Fig. 15 (a) Transmittance change of the ECD-1–4 and ECD-1p–4p under different V_ap at 580 nm under a different V_ap condition (1 : 1.0/−2.0 V, 2 : 1.3/−2.0 V, 3 : 1.5/−2.0 V, 4 : 1.8/−2.0 V, and 5 : 2.0/−2.0 V) (b) charging and discharging cycles for ECD-3p (blue, solid line) and the simultaneous transmittance change (black) up to 300 cycles under 1.5/−2.0 V, and charging/discharging cycles for ECD-3p under 0/1.8 V (red); (c) charging and discharging test of the ECCWs at the current of 0.2 mA (voltage scale from −1.8 V to 1.8 V) for ECDs having different TiO₂(NP)–PW₁₂ thicknesses: 1P = 633 nm, 2P = 845 nm, 3P = 1190 nm, 4P = 1930 nm. (d) Cyclic stability of ECD-3p for V_s of 1.8/−1.8 V; inset: the bistability of ECCW-3p for 3.5 and for 2.35 h after voltage-off from −2.0 and 1.5 V, respectively. (e) Correlation between the PW₁₂ content and the charging/discharging capacitance for ECD-1p–4p.¹⁷³ Reproduced from ref. 173, Copyright 2019, with permission from John Wiley & Sons, Inc.

As described before, an ECD with low-HOMO-level EC polymers and ionic liquid electrolytes can block IET and IDT to provide a bistable ECD. In the presence of PW₁₂, the optical contrast loss was determined as 6% for 3.5 h and 9% for 2.35 h after voltage-off from −2.0 and 1.5 V, respectively, showing improved bistability of the ECD (Fig. 15d, inset).

Concurrently, POMs are good candidates for a pseudocapacitor^180,181 because they function as an electron reservoir. Although the EC polymer in the above-mentioned ECD is not a good capacitor and the electrolyte is an acid-free ionic electrolyte, the capacitance from the above PR-Br/BMIL-TFSI/TiO₂(NP)–PW₁₂ system is notable when subjected to a constant current charge and discharge condition (Fig. 15c, solid line). This can provide ECCW that combines electrochromism and an energy storage system. The capacitance of the ECCW based on PR-Br/BMIL-TFSI/TiO₂(NP)–PW₁₂ increased with the thickness of TiO₂(NP)–PW₁₂ or the content of PW₁₂ (Fig. 15e) until it reached the resistance limit. In addition, the Coulombic efficiency for each complete cycle for these ECCWs was >98%, which was higher than that of the corresponding ECCW without PW₁₂. The charging process was simultaneously accompanied by the color change of ECCW and even after the extended galvanostatic charging–discharging cycles, no optical loss was found (Fig. 16b).

Fig. 16 (a) Ragone plot of various electrochemical capacitors. (b) Figure of merit for color contrast against power density for various EC capacitors.¹⁷³ Reproduced from ref. 173, Copyright 2019, with permission from John Wiley & Sons, Inc.

Because all the materials in this ECCW are not traditional supercapacitor materials, the capacitive performances were lower compared to that of the traditional supercapacitors. However, the thicknesses of the EC materials and redox shuttle (POM) are quite low; thus, the power density of this ECCW was higher than that of various oxides and conjugated polymer systems,^182–189 which often use acidic electrolytes (Fig. 16a). Since the ECCWs with TiO₂(NP)–PW₁₂ use acid-free ionic electrolytes, it can provide a way to develop a non-aqueous EC capacitor with high color contrast and long cyclic stability (Fig. 15b). Fig. 16b shows a high figure-of-merit for color contrast against power density (f_PDT) for several ECCWs. The TiO₂(NP)–PW₁₂ system showed the highest figure of merit for the color contrast among the ECCWs in Fig. 16b. Taking advantage of the high color contrast and power density, it was possible to use the TiO₂(NP)–PW₁₂ system for lightening an LED (Fig. 17) via energy transfer from the charged ECCW to the LED.

Fig. 17 (a) The equivalent circuit used in lighting an LED. (b and c) Photographic images on the color switching of ECCW (ECD-3p) at (b) −2.0 V and (c) 1.8 V; and (d) under bias free condition, at which ECCW transferred energy (discharge) to the LED for lighting upon G1 connection. The photographic image of the LED lit under a dark environment is shown in (e).¹⁷³ Reproduced from ref. 173, Copyright 2019, with permission from John Wiley & Sons, Inc.

8. Conclusion and perspective

POMs consist of a 3-dimensional close-packed metal oxide framework and act as an electron reservoir, giving rise to mixed-valence colored state species without their structural changes. As a unique feature, POMs can be designed at the molecular level. The two basic structures of Keggin and Dawson type POMs can generate several derivatives such as lacunary, complex and substituted structures. Among the huge family of Dawson-type POMs, only α-Dawson and its derivatives are the most widely used structures in the POM EC films. While the color of the POM molecules is affected by the number of the substituted metal, M, or the degree of lacunary structures, the more the lacunary structures of POM, the larger the EC optical contrast. Furthermore, λ_max blue-shifted with the increase in the degree of lacunary structures. Generally, the EC properties of the Dawson structures are better than those of the Keggin structures and the EC properties of the lacunary structures are better than those of the saturated structures. Stabilizers are introduced for the preparation of a stable and transparent POM film. Large organic cations, polycations, carbon materials, and metal oxide semiconductors have been researched as stabilizers and linkers. Therefore, the EC properties of POMs can be improved not only by the structural and atomic variation of POMs but also by the selection of stabilizers and linkers. Among the stabilizers, TiO₂(NPs) have been examined widely and show very high EC optical contrast and long cyclic stability. Other metal oxides and carbon materials have emerged as stabilizers and contribute to finding new functionalities. Over the years, it has been demonstrated that POMs are promising candidates for EC applications. Until now, the TiO₂(NP)–P₂W₁₅V₃-based EC film showed the highest optical contrast of 91.8% in an aqueous electrolyte. The TiO₂(NP)-P₅W₃₀ composite film showed the shortest response time of less than 1 s in a non-aqueous electrolyte with cyclic stability of over 1000 cycles. However, most of the studies on EC performance for POMs have been carried out in a half cell using a three-electrode system. Since the practical application requires a full cell (Type 3), there are still many factors to be understood to develop a highly efficient POM-based ECD. For example, the electrolyte, CE, and device structure should cooperate to afford an ECD with high optical contrast, fast response time, and high coloration efficiency. Furthermore, POMs find new functions in EC devices as a (1) redox shuttle in the charge balancing reaction and (2) capacitive materials. Even with an ultra-thin coating of POM, one can find a drastic change in driving voltage and large enhancement in the energy storage effect. A perspective on possible research directions is also given along with POM materials for new emerging integrative devices.

Abbreviations

P2W18	[P2W18O62]^6−
P2W17	[P2W17O61]^10−
P2W15	[P2W15O56]^12−
P2W12	[H2P2W12O48]^12−
P2W17V	[P2W17VO62]^7−
P2W16V2	[P2W16V2O62]^8−
P2W15V3	[P2W15V3O62]^9−
P2W17Ni	[P2W17O61Ni(H2O)]^8−
As2W15Mg3	[As2W15Mg3O62]^18−
Ce(P2W17)2	[Ce(P2W17O61)2]^17−
Na-P5W30	[NaP5W30O110]^14−
Eu–P5W30	[Eu(OH2)P5W30O110]^12−
S5W30	[NaS5W30O110]^9−
P8W48	[P8W48O184]^40−
PW12	[PW12O40]^3−
PW11V	[PW11VO40]^4−
PW10V2	[PW10V2O40]^5−
PW9V3	[PW9V3O40]^6−
SiW9V3	[SiW9V3O40]^7−
PMo12	[PMo12O40]^3−
ECDs	Electrochromic devices
EC	Electrochromic
POMs	Polyoxometalates
CPs	Conjugated polymers
CE	Counter electrode
LBL	Layer-by-layer
ECCW	EC capacitive Window
EES	EC energy storage
OM	Optical memory
V OFF	Voltage-off state
V ap	applied potential
IET	Interfacial electron transfer
IDT	Interfacial dopant ion transfer
O → M	Oxygen-to-metal
LMCT	Ligand-to-metal charge transfer
η ce	Coloration efficiency
WE	Working electrode
SCE	Saturated calomel electrode
CVD	Chemical vapor deposition
OMPE	Oxymethylene polyoxyethylene
PAH	Poly(allylamine hydrochloride)
PSS	Poly(stryenesulfonate)
PDDA	Poly(diallyl dimethyl ammonium chloride)
PEI	Poly(ethylenimine)
LUMO	Lowest unoccupied molecular orbital
PVA	Poly(vinyl alcohol)
NW	Nanowire
P4VP	Poly(4-vinylpyridine)
Iso-POM	Isopolyoxometalate
NP	Nanoparticles
CNT	Carbon nanotube
SWCNTs	Single-walled CNTs
CS	Chitosan
CBL	Charge balancing layer
CV	Cyclic voltammetry
EDS	Energy-dispersive spectroscopy
BMIM-TFSI	1-Butyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide
PR-Br	Poly(3,3-bis(bromomethyl)-3,4-dihydro-2H-thieno[3,4-b][1,4]dioxepine)s
WE	Working electrode
PANI	Polyaniline

Conflicts of interest

All authors declare no conflict of interest.

Acknowledgements

This research was supported by the Global Research Laboratory (GRL) through the National Research Foundation of Korea (NRF) (no. 2016K1A1A2912753), National Natural Science Foundation of China (no. 21501084) and the Korea Foundation for Advanced Studies’ International Scholar Exchange Fellowship for the academic year of 2017–2018.

References

1. B. Kim, H. Shin, T. Park, H. Lim and E. Kim, Adv. Mater., 2013, 25, 5483–5489.
2. A. A. Argun, P.-H. Aubert, B. C. Thompson, I. Schwendeman, C. L. Gaupp, J. Hwang, N. J. Pinto, D. B. Tanner, A. G. MacDiarmid and J. R. Reynolds, Chem. Mater., 2004, 16, 4401–4412.
3. P. M. Beaujuge and J. R. Reynolds, Chem. Rev., 2010, 110, 268–320.
4. M. Sassi, M. M. Salamone, R. Ruffo, G. E. Patriarca, C. M. Mari, G. A. Pagani, U. Posset and L. Beverina, Adv. Funct. Mater., 2016, 26, 5240–5246.
5. H. Shin, S. Seo, C. Park, J. Na, M. Han and E. Kim, Energy Environ. Sci., 2016, 9, 117–122.
6. S. Heo, J. Kim, G. K. Ong and D. J. Milliron, Nano Lett., 2017, 17, 5756–5761.
7. J. Kim, J. You, B. Kim, T. Park and E. Kim, Adv. Mater., 2011, 23, 4168–4173.
8. H. Shin, Y. Kim, T. Bhuvana, J. Lee, X. Yang, C. Park and E. Kim, ACS Appl. Mater. Interfaces, 2012, 4, 185–191.
9. T. Bhuvana, B. Kim, X. Yang, H. Shin and E. Kim, Angew. Chem., Int. Ed., 2013, 52, 1180–1184.
10. M. Han, M. K. Poduval, H. Shin, N. Tamaoki, T. Park, Y. Kim and E. Kim, Opt. Mater. Express, 2017, 7, 2117–2125.
11. E. Kim and Y. Kim, Mol. Cryst. Liq. Cryst., 2006, 447, 173/[491]–179/[497].
12. E. Kim and S. Jung, Chem. Mater., 2005, 17, 6381–6387.
13. Y. Kim and E. Kim, Curr. Appl. Phys., 2006, 6, e202–e205.
14. J. Baek, Y. Kim and E. Kim, J. Nanosci. Nanotechnol., 2008, 8, 4851–4855.
15. W. Xu, K. Fu and P. W. Bohn, ACS Sens., 2017, 2, 1020–1026.
16. X. Lv, W. Li, M. Ouyang, Y. Zhang, D. S. Wright and C. Zhang, J. Mater. Chem. C, 2017, 5, 12–28.
17. S. I. Cho, W. J. Kwon, S. J. Choi, P. Kim, S. A. Park, J. Kim, S. J. Son, R. Xiao, S. H. Kim and S. B. Lee, Adv. Mater., 2005, 17, 171–175.
18. L. Hong, N. Nakashima, Y. Li, H. Jia and C. Yang, Chem. – Asian J., 2017, 13, 210–216.
19. G. Cai, P. Darmawan, X. Cheng and P. S. Lee, Adv. Energy Mater., 2017, 7, 1602598.
20. H. Li, L. McRae, C. J. Firby, M. Al-Hussein and A. Y. Elezzabi, Nano Energy, 2018, 47, 130–139.
21. Y. Zhong, Z. Chai, Z. Liang, P. Sun, W. Xie, C. Zhao and W. Mai, ACS Appl. Mater. Interfaces, 2017, 9, 34085–34092.
22. X. Xia, Z. Ku, D. Zhou, Y. Zhong, Y. Zhang, Y. Wang, M. J. Huang, J. Tu and H. J. Fan, Mater. Horiz., 2016, 3, 588–595.
23. H. Li, L. McRae, C. J. Firby and A. Y. Elezzabi, Adv. Mater., 2019, 1807065.
24. S. Cong, Y. Tian, Q. Li, Z. Zhao and F. Geng, Adv. Mater., 2014, 26, 4260–4267.
25. T. Yamase, Chem. Rev., 1998, 98, 307–326.
26. S. H. Kawai, S. L. Gilat, R. Ponsinet and J.-M. Lehn, Chem. – Eur. J., 1995, 1, 285–293.
27. D. W. Steuerman, H.-R. Tseng, A. J. Peters, A. H. Flood, J. O. Jeppesen, K. A. Nielsen, J. F. Stoddart and J. R. Heath, Angew. Chem., Int. Ed., 2004, 43, 6486–6491.
28. P. Tehrani, L.-O. Hennerdal, A. L. Dyer, J. R. Reynolds and M. Berggren, J. Mater. Chem., 2009, 19, 1799–1802.
29. Z. Q. Tong, S. K. Liu, X. G. Li, J. P. Zhao and Y. Li, Nanoscale Horiz., 2018, 3, 261–292.
30. Z. B. Li, J. H. Kong, F. K. Wang and C. B. He, J. Mater. Chem. C, 2017, 5, 5283–5298.
31. P. H. Yang, P. Sun and W. J. Mai, Mater. Today, 2016, 19, 394–402.
32. C. X. Zhan, G. Q. Yu, Y. Lu, L. Y. Wang, E. Wujcik and S. Y. Wei, J. Mater. Chem. C, 2017, 5, 1569–1585.
33. A. L. S. Eh, A. W. M. Tan, X. Cheng, S. Magdassi and P. S. Lee, Energy Technol., 2018, 6, 33–45.
34. M. Lahav and M. E. van der Boom, Adv. Mater., 2018, 30, e1706641.
35. H. F. Wang, M. Barrett, B. Duane, J. Gu and F. Zenhausern, Mater. Sci. Eng., B, 2018, 228, 167–174.
36. W. Wu, M. Wang, J. M. Ma, Y. L. Cao and Y. H. Deng, Adv. Electron. Mater., 2018, 4, 1800185.
37. H. J. Yen and G. S. Liou, Polym. Chem., 2018, 9, 3001–3018.
38. M. Ammam, J. Mater. Chem. A, 2013, 1, 6291–6312.
39. T. Pope Michael and A. Müller, Angew. Chem., Int. Ed. Engl., 1991, 30, 34–48.
40. D. E. Katsoulis, Chem. Rev., 1998, 98, 359–388.
41. N. Mizuno and M. Misono, Chem. Rev., 1998, 98, 199–218.
42. R. Neumann, in Polyoxometalate Molecular Science, ed. J. J. Borrás-Almenar, E. Coronado, A. Müller and M. Pope, Springer Netherlands, Dordrecht, 2003, pp. 327–349 DOI:10.1007/978-94-010-0091-8_11.
43. E. Coronado, C. Giménez-Saiz and C. J. Gómez-García, Coord. Chem. Rev., 2005, 249, 1776–1796.
44. B. Hasenknopf, Front. Biosci., 2005, 10, 275–287.
45. M. Vazylyev, D. Sloboda-Rozner, A. Haimov, G. Maayan and R. Neumann, Top. Catal., 2005, 34, 93–99.
46. J. M. Clemente-Juan, E. Coronado and A. Gaita-Ariño, Chem. Soc. Rev., 2012, 41, 7464–7478.
47. T. Yamase, in Biomedical Inorganic Polymers: Bioactivity and Applications of Natural and Synthetic Polymeric Inorganic Molecules, ed. W. E. G. Müller, X. Wang and H. C. Schröder, Springer Berlin Heidelberg, Berlin, Heidelberg, 2013, pp. 65–116 DOI:10.1007/978-3-642-41004-8_4.
48. S. Omwoma, W. Chen, R. Tsunashima and Y.-F. Song, Coord. Chem. Rev., 2014, 258–259, 58–71.
49. S.-M. Wang, L. Liu, W.-L. Chen and E.-B. Wang, Electrochim. Acta, 2013, 113, 240–247.
50. C. Brevard, R. Schimpf, G. Tourne and C. M. Tourne, J. Am. Chem. Soc., 1983, 105, 7059–7063.
51. G. M. Brown, M.-R. Noe-Spirlet, W. R. Busing and H. A. Levy, Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem., 1977, 33, 1038–1046.
52. N. Haraguchi, Y. Okaue, T. Isobe and Y. Matsuda, Inorg. Chem., 1994, 33, 1015–1020.
53. R. Massart, R. Contant, J. M. Fruchart, J. P. Ciabrini and M. Fournier, Inorg. Chem., 1977, 16, 2916–2921.
54. N. H. Nsouli, B. S. Bassil, M. H. Dickman, U. Kortz, B. Keita and L. Nadjo, Inorg. Chem., 2006, 45, 3858–3860.
55. P. J. Domaille, J. Am. Chem. Soc., 1984, 106, 7677–7687.
56. G. A. Tsigdinos and C. J. Hallada, Inorg. Chem., 1968, 7, 437–441.
57. B. Dawson, Acta Crystallogr., 1953, 6, 113–126.
58. R. Contant, W. G. Klemperer and O. Yaghi, Inorg. Synth., 2007, 104–111,  DOI:10.1002/9780470132586.ch18.
59. R. G. Finke, M. W. Droege and P. J. Domaille, Inorg. Chem., 1987, 26, 3886–3896.
60. R. G. Finke, B. Rapko, R. J. Saxton and P. J. Domaille, J. Am. Chem. Soc., 1986, 108, 2947–2960.
61. Z.-M. Zhang, S. Yao, Y.-G. Li, X.-B. Han, Z.-M. Su, Z.-S. Wang and E.-B. Wang, Chem. – Eur. J., 2012, 18, 9184–9188.
62. Y. Jeannin, J. Martin-Frere, D. J. Choi and M. T. Pope, Inorg. Synth., 2007, 115–118,  DOI:10.1002/9780470132586.ch20.
63. N. I. Gumerova and A. Rompel, Nat. Rev. Chem., 2018, 2, 0112.
64. R. J. Mortimer, Electrochim. Acta, 1999, 44, 2971–2981.
65. S.-Y. Kao, C.-W. Kung, H.-W. Chen, C.-W. Hu and K.-C. Ho, Sol. Energy Mater. Sol. Cells, 2016, 145, 61–68.
66. G. W. Jang, C. Chen, R. W. Gumbs, Y. Wei and J. M. Yeh, J. Electrochem. Soc., 1996, 143, 2591–2596.
67. D. M. DeLongchamp, M. Kastantin and P. T. Hammond, Chem. Mater., 2003, 15, 1575–1586.
68. G. A. Sotzing, J. R. Reynolds and P. J. Steel, Chem. Mater., 1996, 8, 882–889.
69. C. M. Amb, A. L. Dyer and J. R. Reynolds, Chem. Mater., 2011, 23, 397–415.
70. S. R. Bathe and P. S. Patil, Sol. Energy Mater. Sol. Cells, 2007, 91, 1097–1101.
71. L. D. Kadam and P. S. Patil, Sol. Energy Mater. Sol. Cells, 2001, 69, 361–369.
72. Z. Yu, X. Jia, J. Du and J. Zhang, Sol. Energy Mater. Sol. Cells, 2000, 64, 55–63.
73. H. Huang, J. Tian, W. K. Zhang, Y. P. Gan, X. Y. Tao, X. H. Xia and J. P. Tu, Electrochim. Acta, 2011, 56, 4281–4286.
74. Y. Kim, H. Shin, M. Han, S. Seo, W. Lee, J. Na, C. Park and E. Kim, Adv. Funct. Mater., 2017, 27, 1701192.
75. B. Tell and F. Wudl, J. Appl. Phys., 1979, 50, 5944–5946.
76. S. Liu, L. Xu, G. Gao, B. Xu and W. Guo, Mater. Chem. Phys., 2009, 116, 88–93.
77. J. Padilla, V. Seshadri, G. A. Sotzing and T. F. Otero, Electrochem. Commun., 2007, 9, 1931–1935.
78. P. M. S. Monk, R. J. Mortimer and D. R. Rosseinsky, Electrochromism: Terminology, Scope, Colouration, 2007, pp. 2–21 DOI:10.1002/9783527615377.ch1.
79. J. Deng, M. Gu and J. Di, Appl. Surf. Sci., 2011, 257, 5903–5907.
80. Y. F. Yuan, X. H. Xia, J. B. Wu, Y. B. Chen, J. L. Yang and S. Y. Guo, Electrochim. Acta, 2011, 56, 1208–1212.
81. C.-C. Liao, Sol. Energy Mater. Sol. Cells, 2012, 99, 26–30.
82. A. J. More, R. S. Patil, D. S. Dalavi, S. S. Mali, C. K. Hong, M. G. Gang, J. H. Kim and P. S. Patil, Mater. Lett., 2014, 134, 298–301.
83. P. Yang, P. Sun, Z. Chai, L. Huang, X. Cai, S. Tan, J. Song and W. Mai, Angew. Chem., Int. Ed., 2014, 53, 11935–11939.
84. L. Ma, Y. Li, X. Yu, Q. Yang and C.-H. Noh, Sol. Energy Mater. Sol. Cells, 2009, 93, 564–570.
85. H. Ling, J. Lu, S. Phua, H. Liu, L. Liu, Y. Huang, D. Mandler, P. S. Lee and X. Lu, J. Mater. Chem. A, 2014, 2, 2708–2717.
86. C. Dulgerbaki, N. Nohut Maslakci, A. I. Komur and A. Uygun Oksuz, Electroanalysis, 2016, 28, 1873–1879.
87. L.-P. Gao, G.-J. Ding, Y.-C. Wang and Y.-L. Yang, Appl. Surf. Sci., 2011, 258, 1184–1191.
88. C. E. Patil, N. L. Tarwal, P. S. Shinde, H. P. Deshmukh and P. S. Patil, J. Phys. D: Appl. Phys., 2008, 42, 025404.
89. L. M. Bertus, A. Enesca and A. Duta, Thin Solid Films, 2012, 520, 4282–4290.
90. C. E. Patil, N. L. Tarwal, P. R. Jadhav, P. S. Shinde, H. P. Deshmukh, M. M. Karanjkar, A. V. Moholkar, M. G. Gang, J. H. Kim and P. S. Patil, Curr. Appl. Phys., 2014, 14, 389–395.
91. H. Kamal, E. K. Elmaghraby, S. A. Ali and K. Abdel-Hady, Thin Solid Films, 2005, 483, 330–339.
92. J. D. Desai, S.-K. Min, K.-D. Jung and O.-S. Joo, Appl. Surf. Sci., 2006, 253, 1781–1786.
93. D. Davazoglou, G. Leveque and A. Donnadieu, Sol. Energy Mater., 1988, 17, 379–390.
94. K. A. Gesheva and T. Ivanova, Chem. Vap. Deposition, 2006, 12, 231–238.
95. K. A. Gesheva, T. M. Ivanova and G. Bodurov, Prog. Org. Coat., 2012, 74, 635–639.
96. D. Gogova, A. Iossifova, T. Ivanova, Z. Dimitrova and K. Gesheva, J. Cryst. Growth, 1999, 198199, 1230–1234.
97. T. Ivanova, K. Gesheva, F. Hamelmann, G. Popkirov, M. Abrashev, M. Ganchev and E. Tzvetkova, Vacuum, 2004, 76, 195–198.
98. T. Ivanova, K. A. Gesheva, G. Popkirov, M. Ganchev and E. Tzvetkova, Mater. Sci. Eng., B, 2005, 119, 232–239.
99. J. P. Lock, J. L. Lutkenhaus, N. S. Zacharia, S. G. Im, P. T. Hammond and K. K. Gleason, Synth. Met., 2007, 157, 894–898.
100. T. Maruyama and S. Arai, Sol. Energy Mater. Sol. Cells, 1993, 30, 257–262.
101. T. Ivanova, K. Gesheva and A. Szekeres, J. Solid State Electrochem., 2002, 7, 21–24.
102. S. Papaefthimiou, G. Leftheriotis and P. Yianoulis, Electrochim. Acta, 2001, 46, 2145–2150.
103. N. Miyata and S. Akiyoshi, J. Appl. Phys., 1985, 58, 1651–1655.
104. L. Ottaviano, A. Pennisi, F. Simone and A. M. Salvi, Opt. Mater., 2004, 27, 307–313.
105. A. Subrahmanyam and A. Karuppasamy, Sol. Energy Mater. Sol. Cells, 2007, 91, 266–274.
106. C. M. Wang, C. Y. Wen, Y. C. Chen, K. S. Kao, D. L. Cheng and C. H. Peng, Integr. Ferroelectr., 2014, 158, 62–68.
107. G. Decher, Science, 1997, 277, 1232–1237.
108. S.-M. Wang, L. Liu, W.-L. Chen, Z.-M. Zhang, Z.-M. Su and E.-B. Wang, J. Mater. Chem. A, 2013, 1, 216–220.
109. A. G. Boulay, G. J. T. Cooper and L. Cronin, Chem. Commun., 2012, 48, 5088–5090.
110. S. Takano, A. Kishimoto, K. Hinokuma and T. Kudo, Solid State Ionics, 1994, 70-71, 636–641.
111. X.-M. Zhang, B.-Z. Shan, Z.-P. Bai, X.-Z. You and C.-Y. Duan, Chem. Mater., 1997, 9, 2687–2689.
112. X. Z. You, B. Z. Shan, X. M. Zhang, S. J. Zheng, Z. P. Bai and J. X. Yang, J. Appl. Electrochem., 1997, 27, 1297–1299.
113. T. He, Y. Ma, Y. Cao, W. Yang and J. Yao, J. Non-Cryst. Solids, 2003, 315, 7–12.
114. F. Boussema, A. J. Gross, F. Hmida, B. Ayed, H. Majdoub, S. Cosnier, A. Maaref and M. Holzinger, Biosens. Bioelectron., 2018, 109, 20–26.
115. D. Jeon, H. Kim, C. Lee, Y. Han, M. Gu, B. S. Kim and J. Ryu, ACS Appl. Mater. Interfaces, 2017, 9, 40151–40161.
116. M. Genovese and K. Lian, ACS Appl. Mater. Interfaces, 2016, 8, 19100–19109.
117. M. Yaqub, S. Imar, F. Laffir, G. Armstrong and T. McCormac, ACS Appl. Mater. Interfaces, 2015, 7, 1046–1056.
118. S. Imar, M. Yaqub, C. Maccato, C. Dickinson, F. Laffir, M. Vagin and T. McCormac, Electrochim. Acta, 2015, 184, 323–330.
119. H. Gu, L. Bi, Y. Fu, N. Wang, S. Liu and Z. Tang, Chem. Sci., 2013, 4, 4371–4377.
120. T. M. Anderson, M. A. Rodriguez, F. Bonhomme, J. N. Bixler, T. M. Alam and M. Nyman, Dalton Trans., 2007, 4517–4522.
121. P. Huang, C. Qin, Z.-M. Su, Y. Xing, X.-L. Wang, K.-Z. Shao, Y.-Q. Lan and E.-B. Wang, J. Am. Chem. Soc., 2012, 134, 14004–14010.
122. M. Nyman, T. M. Alam, F. Bonhomme, M. A. Rodriguez, C. S. Frazer and M. E. Welk, J. Cluster Sci., 2006, 17, 197–219.
123. Z. Sun, L. Xu, W. Guo, B. Xu, S. Liu and F. Li, J. Phys. Chem. C, 2010, 114, 5211–5216.
124. S.-M. Wang, L. Liu, W.-L. Chen, Z.-M. Su, E.-B. Wang and C. Li, Ind. Eng. Chem. Res., 2013, 53, 150–156.
125. G. Gao, L. Xu, W. Wang, Z. Wang, Y. Qiu and E. Wang, Electrochim. Acta, 2005, 50, 1101–1106.
126. F.-Q. Zhang, W. Guan, L.-K. Yan, Y.-T. Zhang, M.-T. Xu, E. Hayfron-Benjamin and Z.-M. Su, Inorg. Chem., 2011, 50, 4967–4977.
127. L. E. Briand, G. T. Baronetti and H. J. Thomas, Appl. Catal., A, 2003, 256, 37–50.
128. G. Gao, L. Xu, W. Wang, W. An, Y. Qiu, Z. Wang and E. Wang, J. Phys. Chem. B, 2005, 109, 8948–8953.
129. S. Liu, L. Xu, G. Gao and B. Xu, Thin Solid Films, 2009, 517, 4668–4672.
130. B. Xu, L. Xu, G. Gao and Y. Jin, Appl. Surf. Sci., 2007, 253, 3190–3195.
131. H. Gu, C. Guo, S. Zhang, L. Bi, T. Li, T. Sun and S. Liu, ACS Nano, 2018, 12, 559–567.
132. L. Liu, S.-M. Wang, C. Li, C.-G. Liu, C.-L. Ma and Z.-B. Han, J. Mater. Chem. C, 2015, 3, 5175–5182.
133. G. Gao, L. Xu, W. Wang, Z. Wang, Y. Qiu and E. Wang, J. Electrochem. Soc., 2005, 152, H102.
134. L.-H. Bi, W.-H. Zhou, J.-G. Jiang and S.-J. Dong, J. Electroanal. Chem., 2008, 624, 269–274.
135. G. Gao, L. Xu, W. Wang, W. An and Y. Qiu, J. Mater. Chem., 2004, 14, 2024–2029.
136. J. Luo, G. F. Jin, F. Zhang, Y. Liu, L. J. Chen, S. Q. Xie and J. W. Zhao, Eur. J. Inorg. Chem., 2018, 143–152.
137. P. Gong, J. Pang, C. Zhai and J. Zhao, Solid State Sci., 2018, 78, 7–15.
138. S. Liu, D. G. Kurth, H. Möhwald and D. Volkmer, Adv. Mater., 2002, 14, 225–228.
139. S. Liu, H. Möhwald, D. Volkmer and D. G. Kurth, Langmuir, 2006, 22, 1949–1951.
140. I. Moriguchi and J. H. Fendler, Chem. Mater., 1998, 10, 2205–2211.
141. Z. Kim, Q.-C. Guo, S.-M. Wang, Q. Wu and Z.-B. Han, Thin Solid Films, 2017, 639, 1–6.
142. B. Chakraborty and I. A. Weinstock, Coord. Chem. Rev., 2019, 382, 85–102.
143. X. An, Q. Tang, H. Lan, H. Liu and J. Qu, Appl. Catal., B, 2019, 244, 407–413.
144. G. R. Bertolini, L. R. Pizzio, A. Kubacka, M. J. Muñoz-Batista and M. Fernández-García, Appl. Catal., B, 2018, 225, 100–109.
145. G. Dong, T. Ye, Y. Yang, L. Sheng, D. Xia, J. Wang, X. Fan and R. Fan, ChemSusChem, 2017, 10, 2218–2225.
146. L. Chen, W. Chen, J. Li, J. Wang and E. Wang, ChemSusChem, 2017, 10, 2945–2954.
147. J. S. Li, X. J. Sang, W. L. Chen, L. C. Zhang, Z. M. Zhu, T. Y. Ma, Z. M. Su and E. B. Wang, ACS Appl. Mater. Interfaces, 2015, 7, 13714–13721.
148. S. M. Lauinger, J. M. Sumliner, Q. Yin, Z. Xu, G. Liang, E. N. Glass, T. Lian and C. L. Hill, Chem. Mater., 2015, 27, 5886–5891.
149. J. M. Sumliner, H. Lv, J. Fielden, Y. V. Geletii and C. L. Hill, Eur. J. Inorg. Chem., 2014, 635–644.
150. L. Li, Q.-y. Wu, Y.-h. Guo and C.-w. Hu, Microporous Mesoporous Mater., 2005, 87, 1–9.
151. S. Liu and X. Qu, Appl. Surf. Sci., 2017, 412, 189–195.
152. S.-M. Wang, L. Liu, Z.-Y. Huang and Z.-B. Han, RSC Adv., 2016, 6, 38782–38789.
153. D. Zhang, Y. Chen, H. Pang, Y. Yu and H. Ma, Electrochim. Acta, 2013, 105, 560–568.
154. S. Liu, J. Mater. Sci.: Mater. Electron., 2016, 27, 11118–11125.
155. X. López and J. M. Poblet, Inorg. Chem., 2004, 43, 6863–6865.
156. W. Guan, L. Yan, Z. Su, S. Liu, M. Zhang and X. Wang, Inorg. Chem., 2005, 44, 100–107.
157. J. M. Maestre, X. Lopez, C. Bo, J.-M. Poblet and N. Casañ-Pastor, J. Am. Chem. Soc., 2001, 123, 3749–3758.
158. S. Usai and J. J. Walsh, J. Electroanal. Chem., 2018, 815, 86–89.
159. C. Li, K. P. O’Halloran, H. Ma and S. Shi, J. Phys. Chem. B, 2009, 113, 8043–8048.
160. H. Tanaka, M. Akai-Kasaya, A. TermehYousefi, L. Hong, L. Fu, H. Tamukoh, D. Tanaka, T. Asai and T. Ogawa, Nat. Commun., 2018, 9, 2693.
161. V. Y. Evtushok, A. N. Suboch, O. Y. Podyacheva, O. A. Stonkus, V. I. Zaikovskii, Y. A. Chesalov, L. S. Kibis and O. A. Kholdeeva, ACS Catal., 2018, 8, 1297–1307.
162. R. Liu, G. Zhang, H. Cao, S. Zhang, Y. Xie, A. Haider, U. Kortz, B. Chen, N. S. Dalal, Y. Zhao, L. Zhi, C.-X. Wu, L.-K. Yan, Z. Su and B. Keita, Energy Environ. Sci., 2016, 9, 1012–1023.
163. N. Kawasaki, H. Wang, R. Nakanishi, S. Hamanaka, R. Kitaura, H. Shinohara, T. Yokoyama, H. Yoshikawa and K. Awaga, Angew. Chem., Int. Ed., 2011, 50, 3471–3474.
164. F. M. Toma, A. Sartorel, M. Iurlo, M. Carraro, P. Parisse, C. Maccato, S. Rapino, B. R. Gonzalez, H. Amenitsch, T. Da Ros, L. Casalis, A. Goldoni, M. Marcaccio, G. Scorrano, G. Scoles, F. Paolucci, M. Prato and M. Bonchio, Nat. Chem., 2010, 2, 826–831.
165. J. Hu, Y. Ji, W. Chen, C. Streb and Y.-F. Song, Energy Environ. Sci., 2016, 9, 1095–1101.
166. A. Ma, X. Zhang, Z. Li, X. Wang, L. Ye and S. Lin, J. Electrochem. Soc., 2014, 161, F1224–F1230.
167. H. Zhang, A. Xie, Y. Shen, L. Qiu and X. Tian, Phys. Chem. Chem. Phys., 2012, 14, 12757–12763.
168. H. Li, S. Pang, S. Wu, X. Feng, K. Mullen and C. Bubeck, J. Am. Chem. Soc., 2011, 133, 9423–9429.
169. L. Huang, J. Hu, Y. Ji, C. Streb and Y.-F. Song, Chem. – Eur. J., 2015, 21, 18799–18804.
170. D. Ma, L. Liang, W. Chen, H. Liu and Y.-F. Song, Adv. Funct. Mater., 2013, 23, 6100–6105.
171. S. Liu, L. Xu, F. Li, W. Guo, Y. Xing and Z. Sun, Electrochim. Acta, 2011, 56, 8156–8162.
172. S. Liu, L. Xu, F. Li, B. Xu and Z. Sun, J. Mater. Chem., 2011, 21, 1946–1952.
173. S. M. Wang, Y. Kim, B. Kim, M. Han and E. Kim, Adv. Funct. Mater., 2019, 29, 1806590.
174. H. Li, J. Gupta, S. Wang, N. Zhang and C. Bubeck, J. Colloid Interface Sci., 2014, 427, 25–28.
175. M. Sadakane and E. Steckhan, Chem. Rev., 1998, 98, 219–238.
176. J. F. Jover, R. Lugo, H. Toulhoat, P. Simon and T. de Bruin, J. Phys. Chem. C, 2014, 118, 864–872.
177. A. Hiskia, A. Mylonas and E. Papaconstantinou, Chem. Soc. Rev., 2001, 30, 62–69.
178. S. Yanagida, A. Nakajima, T. Sasaki, Y. Kameshima and K. Okada, Chem. Mater., 2008, 20, 3757–3764.
179. D. Zhou and B.-H. Han, Adv. Funct. Mater., 2010, 20, 2717–2722.
180. Y. Nishimoto, D. Yokogawa, H. Yoshikawa, K. Awaga and S. Irle, J. Am. Chem. Soc., 2014, 136, 9042–9052.
181. H. Wang, S. Hamanaka, Y. Nishimoto, S. Irle, T. Yokoyama, H. Yoshikawa and K. Awaga, J. Am. Chem. Soc., 2012, 134, 4918–4924.
182. D. Wei, M. R. J. Scherer, C. Bower, P. Andrew, T. Ryhänen and U. Steiner, Nano Lett., 2012, 12, 1857–1862.
183. C.-H. Shan, H. Zhang, W.-L. Chen, Z.-M. Su and E.-B. Wang, J. Mater. Chem. A, 2016, 4, 3297–3303.
184. D. Xu, X. Xiao, J. Cai, J. Zhou and L. Zhang, J. Mater. Chem. A, 2015, 3, 16424–16429.
185. S.-M. Wang, L. Liu, W.-L. Chen, Z.-M. Su, E.-B. Wang and C. Li, Ind. Eng. Chem. Res., 2014, 53, 150–156.
186. B. K. K. Kasem and S. Jones, Platinum Met. Rev., 2008, 52, 100–106.
187. G. Inzelt, A. Lewenstam and F. Scholz, Handbook of reference electrodes, Springer, 2013.
188. L. A. Estrada, D. Y. Liu, D. H. Salazar, A. L. Dyer and J. R. Reynolds, Macromolecules, 2012, 45, 8211–8220.
189. Z.-S. Wu, W. Ren, D.-W. Wang, F. Li, B. Liu and H.-M. Cheng, ACS Nano, 2010, 4, 5835–5842.

---