Electrochemically driven physical properties of solid-state materials: action mechanisms and control schemes

Abstract

The various physical properties recently induced by solid-state electrochemical reactions must be comprehensively understood, and their mechanisms of action should be elucidated. Reversible changes in conductivity, magnetism, and colour have been achieved by combining the redox reactions of d metal ions and organic materials, as well as the molecular and crystal structures of solids. This review describes the electrochemically driven physical properties of conductors, magnetic materials, and electrochromic materials using various electrochemical devices.

---

Takeshi Shimizu

Takeshi Shimizu is an assistant professor at the National Institute of Technology, Yonago College. He received his Ph.D. degree in physics from Kwansei Gakuin University, in 2019. His research interests focus on next-generation cathode materials for rechargeable batteries and the energy storage mechanism of the electrode materials.

Hirofumi Yoshikawa

Hirofumi Yoshikawa is a professor at Kwansei Gakuin University. He received his Ph.D. degree in inorganic chemistry from the University of Tokyo, in 2003. His research interests focus on electrochemistry and rechargeable batteries using organic and inorganic compounds, such as polyoxometalates and metal organic frameworks.

---

1. Introduction

External-stimuli-responsive solid-state materials have recently garnered research attention because of their functions and capability of imparting their properties to useful devices.^1–3 These materials offer three practical advantages (Fig. 1): (1) easily and quickly switchable properties controlled by stimuli such as light,⁴ heat,⁵ magnetic field,⁶ and electricity;⁷ (2) diverse physical properties such as electronic conductivity and magnetism induced by their components (atoms, ions, and molecules) and the inter-component interactions in their crystal structures; and (3) a high-memory effect.⁸ In particular, the electron-associated physical properties can be precisely controlled using a potential or voltage, thereby helping establish new principles for altering physical attributes and measurement systems for monitoring the variations in the physical properties induced by electrochemical reactions. This suggests that the electrochemically responsive solid-state materials are worthy of practical applications; moreover, understanding the relationship between the physical properties and structures of materials at certain potentials or voltages is crucial.

Fig. 1 Advantages of materials responsive to different types of stimuli.

Most electrochemically responsive materials exhibit physical properties based on the characteristics of the redox-active d transition metal ions and organic compounds. The d metal ions, which contain electrons likely to be localised in the d orbitals, form strongly correlated electron systems. This indicates that the d metal ions permit the following four outcomes: first, various physical properties can be determined based on the positions of the d metal ions in the crystal structures, making material design intuitive. Second, the inherent degrees of freedom for charge, spin, and d-orbitals⁹ lead to significant changes in the physical properties under external stimuli. Third, d metal ions exhibit many reversible redox states at certain potentials or voltages,¹⁰ implying that physical property changes can be readily induced by controlling the valences of the d metal ions. Fourth, the d orbitals split depending on the crystal field around the d metal ions,¹¹ forming various band structures and exhibiting physical properties such as electronic conductivity, magnetism, and different chromic effects. Organic compounds have also been used as functional materials because of the following three advantages: (i) previously reported reactions have been found to facilitate the introduction of redox-active sites into the organic parent skeletons, thus simplifying the synthesis-route-related aspects such as the selection of starting materials, reagents, and synthesis method. (ii) Unlike metal oxides, organic molecules are amenable for prediction and improvement of their physical properties through molecular orbital calculations, enabling the efficient synthesis of functional materials. (iii) Organic compounds such as π-conjugated structures can be used as pigments and luminescent materials because their permissible π–π* transitions can be leveraged as light antennas. These transitions enable organic compounds to absorb light of certain wavelengths, and cause complexes with d metal ions¹² or f metal ions¹³ to exhibit electronic transitions, resulting in various pigments and luminescent materials. As mentioned above, electrochemically driven physical properties have been realised using d metal ions and organic compounds. In this context, the present review comprehensively surveys the electrochemical control methods for redox-active functional materials and the various changes in their physical properties, based on the structures of the materials and the valences of the d metal ion and organic molecule components.

2. Electrochemical reaction system

The physical properties of redox-active materials have typically been controlled by solid-state electrochemical reactions; this is a potent approach for inducing redox reactions of materials on electrodes via simultaneous electron transfer and ion intercalation–deintercalation in (1) cell-type or (2) transistor-type devices. In cell-type devices comprising two or three electrodes and an electrolyte, electrons and ions flow through the wire and electrolyte, respectively, in a simultaneous manner at certain potentials (or voltages), causing redox reactions of the materials on the electrodes. These devices include two- and three-electrode systems, with the latter having a reference electrode to determine the standard potential, thereby permitting precise control of the potential and refinement of the physical properties. Two-electrode systems (Fig. 2(b)) have a simple structure with only a working electrode and counter electrode, making it easy to downsize the cells. By combining these downsized cells and a measurement apparatus, changes in various in situ physical properties can be simultaneously controlled and observed. Furthermore, electric double-layer transistors (EDLTs) have mainly been used as transistor-type devices. EDLTs comprise three electrodes (source, drain, and gate electrodes) on an active material and a liquid electrolyte, which forms a high-density electric double layer on the surface of the active material upon the application of a gate voltage V_G (Fig. 2(c)) and induces electrochemical reactions of the active material, such as intercalation of ions in the electrolyte and deintercalation of ions from the active material at the applied overpotential (Fig. 2(d)).¹⁴ These systems can inject more dense carriers than those derived from chemical doping and adjust the ratio of ions in active materials, thereby manifesting various physical properties. As mentioned previously, electrochemical reaction systems have contributed to the discovery of electrochemically induced conductivity, magnetism, and chromism.

Fig. 2 Schematics of (a) a three-electrode system, (b) apparatus equipped with a miniscule cell, and (c and d) an electric double-layer transistor operating (c) below and (d) above the overpotential.

3. Electrochemically driven electronic conductivity and superconductivity

3.1. Mechanism of electronic conduction

A carrier migrates through vacant orbitals in molecular or crystal structures, and its conductivity σ is expressed as

σ = neμ,

which suggests that σ is proportional to the carrier concentration n, elementary charge e, and mobility μ.¹⁵μ reflects three types of carrier conduction mechanisms: (1) band conduction,¹⁶ (2) hopping conduction,¹⁶ and (3) carrier doping.¹⁷ In band conduction (or intramolecular conduction), carriers move within molecules or crystals via covalent bonds, and their conductivities tend to decrease at high temperatures. When valence and conduction bands are formed by infinite covalent bonds between atoms, carriers move in the partially filled band (Fig. 3(a)). To realise this conduction mechanism, components with unpaired electrons such as Cu²⁺ (d⁹) and radicals¹⁸ have been introduced. However, band conduction degrades at high temperatures because the thermal vibrations of atoms interfere with carrier movement. In hopping conduction (or intermolecular conduction), carriers localised at a specific site in a molecule or crystal move to an adjacent site, and their conductivities tend to increase at high temperatures, in contrast to that in band conduction. Carriers with more energy than the energy gap E_g can hop toward higher discretised levels; moreover, hopping conduction is accelerated by the addition of thermal energy (Fig. 3(b)). Carrier doping has been utilised not only for elemental substitution in inorganic compounds but also for redox reactions in conductive polymers.¹⁷ In particular, one-dimensional (1D) charge-injected conducting polymers exhibit metal-type electronic conductivity via band conduction and hopping conduction through intermolecular chains and between intramolecular chains, respectively, although electrochemically neutral 1D conducting polymers with π-conjugated systems, such as polyacetylene and polythiophene, are insulators or semiconductors. Undoped 1D conductive polymers are insulating or semiconducting in nature because the Peierls transition creates an energy gap E_g between the valence and conduction bands. After doping a conducting polymer with the carrier, its conductivity can be improved owing to simultaneous changes in the molecular and band structures. Carrier doping can be achieved in two types of molecular structures: (1) degenerate and (2) non-degenerate systems. Degenerate systems have a symmetrical structure similar to that of polyacetylene, and the energy after bond alternation is equal to that before bond alternation. In this case, a neutral soliton – a bond defect that connects the A and B phases with swapped bond alternations – can stably exist (Fig. 3(c)). The neutral soliton can freely diffuse regardless of its position because the energy of phase A is equal to that of phase B, and the solitons when oxidised or reduced (via p-doping and n-doping, respectively) act as carriers, positive solitons, and negative solitons. Carrier doping can alter the band structures of conducting polymers by the extraction (p-doping) or injection (n-doping) of electrons from the highest occupied molecular orbital (HOMO) or into the lowest unoccupied molecular orbital (LUMO) (Fig. 3(d)). Specifically, carrier doping extracts or injects electrons from or into neutral soliton orbitals, which are derived from neutral degenerate conducting polymers with swapped bond alternations, and formed between the HOMO and LUMO because solitons do not participate in covalent bonding. A high dopant concentration changes the soliton orbitals to wide soliton bands, enhancing the metallic conductivity of the degenerate conducting polymer. Since the discovery of its electrical properties by Shirakawa et al.,¹⁹ polyacetylene has been investigated as an electrochemically doped polymer²⁰ and anode active material for lithium-ion batteries.²¹ Non-degenerate systems such as polyparaphenylene, polypyrrole, and polyaniline have A and B phases with swapped bond alternations of different energies. Typically, benzonoid and quinoid structures are considered as the A and B phases. In these systems, when electrons are extracted from the double bond, energy is required to convert the benzonoid structure into the quinoid configuration. Therefore, quinoid-structured polarons and bipolarons become carriers only in a limited region under high-concentration doping (Fig. 3(e)). This corresponds to the extraction of electrons from neutral non-degenerate conducting polymers (Fig. 3(f)), resulting in higher and lower orbitals between the HOMO and LUMO. Consequently, bipolaron bands are formed under high-concentration doping, improving the metallic conductivity of the non-degenerated conducting polymers.

Fig. 3 (a and b) Schematics of electronic conduction¹⁶ achieved via (a) band conduction and (b) hopping conduction. (c–f) Change in potential energy for degenerate and non-degenerate conjugated polymers:¹⁷ (c) potential energy of phases A and B formed in the degenerate conjugated polymer; (d) energy band variations of the degenerate conjugated polymer due to the creation of a neutral polymer, negative soliton, positive soliton, and soliton band; (e) potential energy of aromatic and quinoid structures in the non-degenerate conjugated polymer; (f) energy band variations of the degenerate conjugated polymer due to the creation of a polaron, bipolaron, and bipolaron band. (a and b) Reproduced from ref. 16 with permission. Copyright 2021, American Chemical Society. (c–f): Reproduced from ref. 17 with permission. Copyright 2020, Elsevier.

As mentioned before, electronic conductivity can be induced by electrochemical redox reactions. Moreover, redox-active complexes and metal oxides have been investigated in addition to conducting polymers, and a correlation between the molecular or crystal structure and conductivity has been established. This section discusses recent research on the mechanisms governing electrochemically driven electronic conductivity based on the molecular or crystal structures and redox states of materials.

3.2. Electrochemically altered electronic conductivity

3.2.1 Conductive polymers. To our knowledge, conductive polymers are considered preeminent materials for modification of electronic conductivity by electrochemical redox reactions. Notably, poly(1-vinylpyrene) – first reported in 1986²² – is a p-dopable non-conjugated conductive polymer that undergoes a redox reaction at 0.71 V vs. Ag/Ag⁺ 0.01 mol dm⁻³ (Fig. 4(a)) in a three-electrode cell containing a Bu₄NClO₄ solution, with ClO₄⁻ acting as a dopant, leading to a change in electronic conductivity. The conduction mechanism of poly(1-vinylpyrene) evidently involves electron hopping between the π-conjugated pendant pyrene, because the pressed polymer exhibits high electronic conductivity. Additionally, the change in electronic conductivity differs between poly(1-vinylpyrene) deposited on the electrode via an electrochemical redox reaction in a polymer solution (electronic conductivity: 8.8 × 10⁻⁸ S cm⁻¹, doping degree: 35%) and its counterpart placed on the electrode (electronic conductivity: 2.6 × 10⁻¹⁰ S cm⁻¹, doping degree: 11%). However, the redox reaction is irreversible owing to the coupling reaction between pendant perylene radical cations in a potential range of 0.42–0.48 V vs. 0.01 mol dm⁻³ Ag/Ag⁺. This reaction has also been observed in a three-electrode cell with a poly(3-vinylperylene) electrode (Fig. 4(b)).²³

Fig. 4 Doping mechanisms of the non-conjugated polymers (a) poly(1-vinylpyrene) and (b) poly(3-vinylperylene).

Most conjugated conductive polymers undergo reversible redox reactions. These systems have skeletons of p-type conductive polymers such as polypyrrole (PPy)²⁴ (Fig. 5(a)), polythiophene (PT)²⁵ (Fig. 5(a)), and polyaniline (PANI)²⁶ (Fig. 5(b)), thereby exhibiting an electronic conductivity change based on the redox reaction accompanied by aromatic–quinoid non-degeneracy. Recently, the electronic conductivity of poly(3-hexylthiophene) (P3HT) was changed from 1 × 10⁻¹ (3.5 V vs. Li⁺/Li) and 7 × 10⁻² (4.0 V vs. Li⁺/Li) to 8 × 10⁻⁶ S cm⁻¹ in a three-electrode cell with a LiTFSI-containing electrolyte and lithium-metal-based counter and reference electrodes, despite maintaining the conductivity of poly(3,4-ethylenedioxythiophene):poly(4-styrene sulfonate) (PEDOT:PSS) at 1 S cm⁻¹.²⁷ Additionally, P3HT–polyethylene oxide (PEO) showed a variation in electronic conductivity from 10⁻⁷ S cm⁻¹ to 10⁻⁴–10⁻² S cm⁻¹ (3.2–3.6 V, 90 °C) in a two-electrode cell with a LiTFSI-containing electrolyte and lithium metal anode.²⁸ Furthermore, the n-type conductive polymer poly{[N,N′-bis(2-octyldodecyl)-naphthalene-1,4,5,8-bis-(dicarboximide)-2,6-diyl]-alt-5,5′-(2,2′-bithiophene)} (P(NDI2OD-T2)) also showed a change in electronic conductivity from 5 × 10⁻⁴ (<2.2 V vs. Li⁺/Li) to 5 × 10⁻⁷ S cm⁻¹ in the three-electrode cell (Fig. 5(c)).²⁷

Fig. 5 Redox reactions of the p-type conjugated polymers (a) polypyrrole (X = NH), polythiophene (X = S), and (b) polyaniline. (c) Redox reaction of the n-type conjugated polymer P(NDI2OD-T2).

3.2.2 Metal–organic frameworks (MOFs). Metal–organic frameworks (MOFs), which are porous materials comprising metal ions and organic ligands, have attracted considerable attention for diverse applications owing to their redox activities.²⁹ In particular, conductive MOFs have been targeted to study synthesis methods and achieve switchable redox-chemistry-based electronic conductivity.^30–36 Recently, Naph(COOLi)₂ was reported as the first MOF with electrochemically switchable electronic conductivity.³⁷ Naph(COOLi)₂, which comprises lithium ions and 2,6-naphthalenedicarboxylic ligands, has been used as a novel anode active material exhibiting complete electrochemical reversibility despite its insulating nature.³⁸ Essentially, lithium ions are intercalated into the pores near the ligands, and a redox reaction of its ligands is induced, enhancing the electronic conductivity of Naph(COOLi)₂ from 10⁻⁹ to 10⁻⁵ S cm⁻¹ (Fig. 6(a)). According to an ab initio simulation (Fig. 6(b and c)), the bandgap of the reduced state (0.99 eV) is smaller than that of the oxidised state (3.27 eV) because of the electron density distribution between the 2,6-naphthalenedicarboxylic ligands, causing anisotropic electron hopping through the two-dimensional (2D) π-stacking naphthalene layers (Fig. 6(d–g)). However, the electronic conductivity decreases at temperatures above 100 °C despite the action mechanism being hopping conduction, because the crystal structure of Naph(COOLi)₂ is destroyed at high temperatures.

Fig. 6 (a) Schematic of the mechanism governing the redox behaviour of 2,6-Naph(COOLi)₂. Calculated electronic band structures for (b) pristine and (c) Li-intercalated 2,6-Naph(COOLi)₂. 3d electron density distribution of the conduction band minimum (CBM) for (d) pristine and (e) Li-intercalated 2,6-Naph(COOLi)₂ (orange, intercalated Li; green, pristine Li; red, O; blue, C; and white, H). 2d electron density distribution of the CBM across the p-stacking naphthalene layer via the C1, C1′, C2, and C2′ carbons for (f) pristine and (g) Li-intercalated 2,6-Naph(COOLi)₂. (a–g) Reproduced from ref. 37 with permission. Copyright 2017, American Association for the Advancement of Science (AAAS).

The electrochemically switchable electronic conductivity has also been evaluated in terms of the diffusion coefficient for charge transfer – D_hopping. NU-1000 (ligand: carboxy-terminated tetraphenylpyrene, TPPy(COO⁻)₄)³⁹ and MOF-525 (ligand: meso-tetrakis(4-carboxyphenyl) porphyrin)^40–42 have also been reported as MOFs with electrochemically switchable electron-hopping-based conductivities (Fig. 7(a) and (b)). NU-1000 and MOF-525 deposited on the electrodes exhibit reversible changes in D_hoppingvia the electrochemical redox reactions of their ligands. In the case of NU-1000 (Fig. 7(c)), D_hopping varies between 2 × 10⁻¹⁰ and 3 × 10⁻¹¹ cm² s⁻¹ (forward and backward steps, respectively; Fig. 7(d)).³⁹ Moreover, MOF-525 (Fig. 7(e)) shows D_hopping values of 5 × 10⁻¹⁴ (0–1.5 V) and 10 × 10⁻¹⁴ (1.5–0 V) cm² s⁻¹ (Fig. 7(f)).^41,42 The electronic state of the ligand is a crucial factor in the process of electrochemically switching the electronic conductivity of MOFs. Additionally, MOFs are promising switchable electronic conductors that operate via redox reactions.³⁷

Fig. 7 Schematics of microcrystallites with (a) a hexagonal face or (b) an end-on orientation in contact with the electrode. (c) Schematics illustrating the crystal structure of NU-1000 and directional charge transport within it (labelled n₁, n₂, and n₃). (d) Cottrell plots for electrode-supported thin films of electrophoretically deposited NU-1000.³⁹ (a–d) Reproduced from ref. 39 with permission. Copyright 2019, American Chemical Society.

3.3. Superconducting mechanisms

In the superconducting state, a pair of electrons – called a Cooper pair – conducts without the loss of kinetic energy, whereas normally conductive electrons lose their kinetic energy via coulombic interactions between the electrons and cations in the crystal lattices. Superconductors have been studied to elucidate their properties and advance their development based mainly on the (1) Bardeen–Cooper–Schrieffer (BCS) theory and (2) carrier doping into cuprate superconductors.⁴³ In BCS-type superconductors, the first electron is trapped in the locally shrunken lattice owing to coulombic interactions; however, the velocity of the second electron is accelerated by the shrunken lattice with a high positive charge density, indicating that the Cooper pair migrates without the loss of net kinetic energy (Fig. 8(a and b)). In terms of cuprate superconductors, most of them have perovskite structures containing CuO₂ and block layers. Superconductivity can be induced when an electron or hole is doped into the CuO₂ layers by substituting high-valence cations with low-valence equivalents or removing the oxide ion (O²⁻) in the block layers. For example, YBa₂Cu₃O₇ (average valence of copper: 3+) exhibits superconductivity at a superconducting transition temperature T_c of 90 K, although YBa₂Cu₃O₆ containing oxygen vacancies (average valence of copper: 2+) is an insulator.^44–47 This indicates that the superconducting state can be controlled by the electrochemical reaction of copper ions or the intercalation–deintercalation of oxide ions.

Fig. 8 (a and b) Schematics of the superconducting mechanism based on the Bardeen–Cooper–Schrieffer theory.

3.4. Electrochemically induced change in superconductivity

LiTi₂O₄ has been investigated as a BCS-type oxide superconductor with a T_c value of ∼13 K (ref. 48–51) and electrochemical redox activity.⁵² LiTi₂O₄ forms a spinel structure comprising Ti³⁺, Ti⁴⁺, and O²⁻ (Fig. 9(a)), and the number of Li⁺ guest ions in LiTi₂O₄ determines the superconductor–insulator transition (SIT). This suggests that LiTi₂O₄ can transform into a superconductor or insulator by lithiation or delithiation accompanied by the electrochemical Ti³⁺/Ti⁴⁺ redox reaction. In 2010, the superconductivity of LiTi₂O₄ was controlled using an electrochemical redox reaction for the first time.⁵³ During lithiation at 1.3 V vs. Li⁺/Li, the magnetic susceptibility of Li_1+xTi₂O₄ (0.2 < x < 1) decreased to zero as LiTi₂O₄ transitioned to the Li₂Ti₂O₄ phase. Moreover, during delithiation at 4.6 V vs. Li⁺/Li, the T_c value of Li_1+xTi₂O₄ (−0.3 < x < 0) increased from 12 to 13.3 K, although the spinel structure partially degraded and the magnetic susceptibility decreased to zero. In 2015, the electrochemically guided superconductivity switching of LiTi₂O₄ was realised utilising a three-electrode system with a LiCoO₂ counter electrode and a AgCl reference electrode (Fig. 9(b)).⁵⁴ Furthermore, the superconductivity of LiTi₂O₄ was controlled by hole doping (delithiation) and electron doping (lithiation) in the EDLT system under the overpotential. The T_c value of Li_1+xTi₂O₄ repeatedly decreased and increased owing to delithiation and lithiation, respectively, under the following V_G control scheme: 0 → −1.0 → 0 → 2.0 → −2.0 → 0 V (Fig. 9(c)), leading to a final change in T_c from 10.0 to 8.5 K. Similarly, T_c changed repeatedly between 8.5 and 9.5 K upon lithiation and delithiation under the following V_G control scheme: 0 → 2.0 → 0 → 3.5 → 0 V.⁵⁵ However, Li_1±xTi₂O_4−δ showed irreversible SIT in the hole-doping region (V_G, negative; Li⁺, deintercalation; and O²⁻, intercalation) and reversible SIT in the electron-doping region (V_G, positive; Li⁺, intercalation; O²⁻, deintercalation). In the former region, Ti ions in Li_1±xTi₂O_4−δ lost all valence electrons, resulting in an irreversible phase transition (Fig. 9(d)). In the latter region, the lattice parameter reversibly changed between 8.37 and 8.35 Å (Fig. 9(e)), and the underlying mechanism was interpreted based on the enhanced 3d electron correlation.⁵⁶

Fig. 9 (a) Crystal structure of LiTi₂O₄ (atom notations: Li, green; O, red; and Ti, blue⁵¹). (b) Temperature dependence of the magnetic susceptibility for Li1 + xTi₂O₄ obtained by Li intercalation and deintercalation in a three-electrode system.⁵⁴ (c) Reversible superconductive transition temperature of LiTi₂O₄ induced by a redox reaction proceeding under the overpotential in an EDLT.⁵⁵In situ powder X-ray diffractometry patterns of LiTi₂O₄ during (d) hole doping and (e) electron doping.⁵⁶ (a) Reproduced from ref. 51 with permission. Copyright 2022, Elsevier. (b): Reproduced from ref. 54 with permission. Copyright 2015, Springer Nature. (c) Reproduced from ref. 55 with permission. Copyright 2015, AIP Publishing LLc. (d and e) Reproduced from ref. 56 with permission. Copyright 2021, American Physical Society.

Electrochemical control over the superconductivity of electron-doped cuprate superconductors such as Li_xSr₂CuO₂Br₂ (T_c = 8.0 K, 0.05 < x < 0.34; V = 0.5 V vs. Li⁺/Li) and Li_xSr₂CuO₂I₂ (T_c = 4.5 K, x = 0.25; V = 0.5 V vs. Li⁺/Li)⁵⁷ has also been investigated because lithiation can alter T_c. However, most hole-doped or non-porous cuprate superconductors lose their superconductivity during lithiation because of lattice deformation; therefore, superconductivity switching cannot be realised using two- or three-electrode systems. In 2018, YBa₂Cu₃O_7−δ (Fig. 10(a)) was found to exhibit reversible SIT in an electrochemical oxygen-doped EDLT system (Fig. 10(b)).⁵⁸ The YBa₂Cu₃O_7−δ insulator transformed into a superconductor as oxide doping proceeded (Fig. 10(c and d)). Furthermore, the insulators KTaO₃ ⁵⁹ and SrTiO₃ ⁶⁰ showed superconductivity at 50 mK (at V_G = 4.5 V) and 0.4 K (at V_G = 3.5 V), respectively, in an EDLT system. Therefore, EDLT systems can aid in exploring new superconductors and inducing superconductivity.

Fig. 10 (a) Crystal structure of YBa₂Cu₃O_7−δ (atom notations: Y, grey; Ba, green; Cu, blue; Y, blue; and O, red⁴⁷). (b) Schematic of lateral oxygen diffusion in a device (red spheres: oxygen ions).⁵⁸ (c) Reversible superconductor–insulator transition of YBa₂Cu₃O_7−δ after applying different voltage pulses during oxygen doping in an EDLT system.⁵⁸ (d) Frequency of the O(4) Ag yttrium barium copper oxide (YBCO) Raman mode obtained at different points along a 50 μm-long 50 μm-wide uniform track patterned in a YBCO film grown on a LAALO3 substrate, in the pristine state (solid symbols) and after switching the left contact in the high-resistance state (open symbols) (inset: schematic of oxygen diffusion within the track).⁵⁸ (a) Reproduced from ref. 46 with permission. Copyright 2019, Springer Open. (b–d) Reproduced from ref. 58 with permission. Copyright 2018, American Chemical Society.

To date, the electrochemical control of electronic conductivity and superconductivity has been scarcely reported. However, molecular and crystal structures have been related to the mechanism governing electrochemically switchable electronic conduction. Moreover, superconductivity has been altered via electrochemical reactions suitable for electron- and hole-doped superconductors. Thus, studies on the electrochemical control of electronic conductivity and superconductivity have facilitated the development of electrochemical switching devices.

4. Electrochemically driven magnetisation

4.1 Principle of magnetisation

Recently, magnetism was found to be controlled by solid-state electrochemical redox reactions. The magnetic behaviour is determined by a combination of the interactions between local unpaired electron spins (up spin: ↑, down spin: ↓) in the metal ions or organic radicals in crystal structures (Fig. 11).⁶¹ Paramagnetic behaviour, in which the spins point in the same direction as the external magnetic field, is exhibited if the spins do not interact with each other in the material. Ferromagnetic behaviour is induced by interactions between the spins with the same moment and directionality (↑↑ or ↓↓) in the material even in the absence of an external magnetic field below the Curie temperature T_c. Antiferromagnetic behaviour is induced by interactions between spins with the same moment and opposite directionality (↑↓ or ↓↑) in the absence of an external magnetic field below the Néel temperature T_N. However, ferrimagnetism arises in spinel structures because spins with different antiparallel magnetic moments interact with each other, resulting in a net magnetic moment. Essentially, magnetism depends on the manner in which the electron spins interact with each other in crystal structures, suggesting that magnetism can be modified by the addition/removal of electrons to/from metal ions or organic radicals. In recently reported studies, solid-state electrochemical redox reactions have induced various magnetisation phenomena at specific voltages. In this section, voltage-driven magnetisation is reviewed considering the changes in the crystal structures and the interactions between the electron spins.

Fig. 11 Schematics of spin interactions resulting in magnetic behaviours. (a) Paramagnetism. (b) Ferromagnetism. (c) Antiferromagnetism. (d) Ferrimagnetism.

4.2. Changes in magnetism induced by redox reactions of metal ions in cell-type electrochemical devices

Electrochemical modulation of magnetism and the underlying mechanism were first reported for chromium-containing materials using cell-type electrochemical devices. Among these systems, CrO₂ – a half-metallic ferromagnetic oxide⁶² – exhibits the simplest change in magnetisation. CrO₂ has a tetragonal rutile structure in which half of the octahedral sites are occupied by Cr⁴⁺ (d²) ions, which have one located electron in the d_xy orbital and an itinerant electron, and an empty tetrahedral site, lifting the degeneracy of the t_2g states owing to tetragonal symmetry. The spins of the itinerant electrons are coupled in parallel between the neighbouring Cr⁴⁺ ions, thereby inducing ferromagnetism.⁶³ Sivakumar et al. reported that the variation in magnetisation was induced by the change in valence from Cr⁴⁺ to Cr³⁺ owing to electrochemical Li insertion into the CrO₂ structure at room temperature (20 °C).⁶² Slight lithiation of Li_xCrO₂ (x = 0.1) led to an 80% decrease in the coercive force (Fig. 12(a)) and 70% saturation magnetisation. This phenomenon was caused by the reduction of Cr⁴⁺ to Cr³⁺, which created d²–d³ and d³–d³ nearest-neighbour pairs and led to the antiparallel coupling of spins through super-exchange interactions between the t_2g orbitals. Further lithiation made the d³–d³ pair more dominant, thereby helping trigger antiferromagnetism and approach zero net magnetisation. However, the magnetism loss of 15%–20% was irreversible despite the reversible electrochemical cycling of specimens with x < 0.1 because of structural degradation or amorphisation during the discharge–charge cycles. Since 1996, redox-active transition metal ions such as Cr and Fe have been used in porous coordination materials to thoroughly monitor the electrochemically driven changes in magnetism. Prussian blue (PB) and Prussian blue analogues (PBAs), which are porous coordination materials comprising transition metal ions and CN⁻ linkers, have a cubic structure with a 5 Å-sized microchannel (M–CN–M′ distance = ∼5 Å);⁶⁴ consequently, they exhibit spin–spin interactions between the transition metal ions⁶⁶ and permit the reversible absorption–desorption of small molecules and ions.⁶⁵ These properties enable PB and PBA systems to undergo electrochemically driven changes in magnetism without experiencing structural decay. Cr_2.43(CN)₆ (Cr^II_1.29Cr^III_1.14(CN)₆) and Cr_2.12(CN)₆ (Cr^II_0.36Cr^III_1.76(CN)₆) (Fig. 12(d)) were the first reported PBAs to exhibit electrochemically switchable magnetism.⁶⁷ The reduction reaction accompanied by K insertion evidently changed the valence from Cr³⁺ to Cr²⁺ at −0.95 V vs. the saturated calomel electrode (SCE) in a 1 M KCl aqueous solution, resulting in a transformation from paramagnetism to ferrimagnetism. Furthermore, the T_c values of Cr^II_1.29Cr^III_1.14(CN)₆ (240 K) and Cr^II_0.36Cr^III_1.76(CN)₆ (270 K) shifted to 120 and 140 K, respectively, indicating that the electrochemical reaction could alter the magnetisation and T_c. In 2013, Yamada et al. reported the detailed magnetic behaviour of Cr^II_0.36Cr^III_1.76(CN)₆.⁶⁸ The specimen showed a reversible change in magnetism between ferrimagnetism (T_c = 270 K) and paramagnetism (T_c = 150 K), based on the change in the spin–spin interactions between the Cr ions occupying the A and B sites of PBA. The seamless change in the magnetism of Cr^II_1.91Cr^III_0.33[Cr^III(CN)₆]·χ1.24Cl_0.13(OH)_1.68·5.25H₂O was monitored using in situ magnetism measurements (Fig. 12(a)). The magnetism of PBA gradually changed with progression of discharge, owing to the two-step reduction of Cr³⁺ to Cr²⁺ between 3.5 and 1.7 V (Fig. 12(b) and (c)). This change in magnetisation was ascribed to the variation in total spin (ST) based on the reduction of Cr³⁺ to Cr²⁺ at the B sites. At 2.2 V (Cr^II_2.24[Cr^III(CN)₆]), the reduction of A-site Cr³⁺ ions increased as T_c increased from 215 to 230 K, because of the change in A-site spin (S_A) from a mix of 2 (Cr²⁺) and 3/2 (Cr²⁺ and Cr³⁺) to 2 (Cr²⁺) (Fig. 12(d)). At 1.7 V (Cr^II_2.24[Cr^II(CN)₆]), the reduction of B-site Cr³⁺ ions enhanced the magnetisation and lowered T_c to ∼160 K. Furthermore, the magnetisation and T_c values of the PBAs were calculated from the spins of Cr at the A and B sites, S_A, S_B, and the associated magnetic coupling constant J, according to mean-field theory.

Fig. 12 (a) Schematic of a lithium battery equipped with an in situ magnetic measurement system.⁶⁸ (b) First discharge curve of the Cr-Prussian blue analogue (PBA) battery.⁶⁸ (c) Temperature dependence of magnetisation for Cr-PBA in the lithium battery equipped with the in situ magnetic measurement system.⁶⁸ (d) Mechanism governing the change in the electrochemical magnetism of Cr-PBA between 3.5 and 1.7 V.⁶⁸ reproduced from ref. 68 with permission. Copyright 2013, Wiley-VCH.

Three bimetallic PBAs and PB specimens featuring other transition metals have also been investigated. First, the simplest ferromagnetic PB specimen, Fe^{III−low spin (LS)}–CN–Ni^II (Ni^II–Fe^III PB, K_0.4Ni^II_1.3[Fe^II(CN)₆]) was used as an electrochemically switchable ferromagnet induced by the Fe^III−LS/Fe^II−LS redox reaction.⁶⁹ In the original state, the redox-inactive Ni^II ion (S_Fe = 1) magnetically interacts with the redox-active Fe^III−LS ion (S_Fe = 1/2), resulting in ferromagnetism. However, in the reduced state, the Ni^II ion (S_Fe = 1) magnetically interacts with the Fe^II−LS ion (S_Fe = 0), leading to paramagnetism. Notably, although the original Ni^II–Fe^III PB specimen exhibited ferromagnetism at T_c < 25 K, paramagnetism was observed at ∼0.4 V vs. SCE. Second, the Fe^III–CN–Cu^II PBA (Li_xCuFe-PBA, 0 < x < 1.75), which exhibited a ferromagnetic transition with a spin configuration (S_Fe = 1/2 and S_Cu = 1/2), was studied as a porous magnetic host.⁷⁰ A change in magnetism was observed for Li_x(CuFe-PBA, 0 < x < 1); however, Fe^III and Cu^II were reversibly reduced to Fe^II and Cu^I, respectively, at the equilibrium voltage (3.5–3.0 V vs. Li/Li⁺) to prevent the gradual phase transition between Li-rich (tetragonal) and Li-poor (cubic) phases at x = 1.34. Pristine CuFe-PBA exhibited ferromagnetic interactions between the Cu^II and Fe^III ions (S_Cu = S_Fe = ½) at temperatures below 20 K. However, after the reduction of Fe^III (S_Fe = 1/2) to Fe^II (S_Fe = 0), Li₁CuFe-PBA showed paramagnetic behaviour over 2–40 K, indicating the elimination of ferromagnetic interactions between the Fe and Cu ions (blue inverted triangles). After the oxidation of Fe^II (S_Fe = 0) to Fe^III (S_Fe = 1/2), Li₀CuFe-PBA exhibited ferromagnetism based on the original spin interactions between the Cu^II (S_Cu = 1/2) and Fe^III (S_Fe = 1/2) ions (red inverted triangles). Moreover, as the ratio of θ and T_c approached unity, the magnetic interactions from second-nearest neighbours (Fe^III–Fe^III and Cu^II–Cu^II) were negligible according to molecular field theory. Third, the core–shell PBA comprising K_0.1Cu[Fe(CN)₆]_0.7·3.5H₂O and K_0.1Ni[Fe(CN)₆]_0.7·4.4H₂O (Li_x(CuFe-PBA@NiFe-PBA)) exhibited complex magnetic behaviour owing to the electrochemical redox reaction of the CuFe-PBA core and NiFe-PBA shell (Fig. 13(a)).⁷¹ A change in magnetisation was observed owing to the reduction of Fe^III/Fe^II and Cu^II/Cu^I in the CuFe-PBA core accompanied by the Fe^III/Fe^II redox reaction in the NiFe-PBA shell (0.35 < x ≤ 0.7) (Fig. 13(b)). Specifically, the original Li_x(CuFe-PBA@NiFe-PBA) (x = 0) exhibited the ferromagnetic transitions derived from CuFe-PBA (T_c = 18 K) and NiFe-PBA (T_c = 24 K) (Fig. 13(c)). For 0 < x ≤ 0.32, the magnetic transition at 18 K gradually disappeared as more Li⁺ ions were intercalated into the CuFe-PBA core, although the ferromagnetic transition derived from NiFe-PBA occurred almost without changes in magnitude and T_c (24 K). For 0.32 < x ≤ 0.64, the magnetisation and T_c (24 K) derived from NiFe-PBA were gradually lowered as substantially more Li⁺ ions were intercalated into the NiFe-PBA shell (Fig. 13(d)), indicating that the electrochemical reaction could separately control the magnetic behaviours of the core and shell structures of PBAs based on the redox reactions of transition metals.

Fig. 13 (a) Schematics illustrating the electrochemical lithiation of the core–shell CuFe-PBA@NiFe-PBA heterostructure.⁷¹ (b) Temperature dependences of field-cooled magnetisation for Li_x(CuFe-PBA@NiFe-PBA) at x = 0.00, 0.11, 0.21, 0.32, 0.42, 0.53, 0.64, 0.74, 0.85, and 1.00.⁷¹ (c and d) Derivatives of field-cooled magnetisation curves for CuFe-PBA (∼18 K at x = 0) and NiFe-PBA (∼24 K at x = 0) upon lithiation.⁷¹ Reproduced from ref. 71 with permission. Copyright 2015, American Chemical Society.

The number of oxides undergoing electrochemically driven magnetisation changes has increased because oxides have different types of crystal structures and metal ions. In particular, normal- or inverse-spinel-structured [M³⁺][M²⁺]O₄ – comprising transition metal ions (M²⁺ and M³⁺) and O²⁻ – has been investigated. The normal spinel structure consists of M³⁺ ions located at tetrahedral sites (A sites) and M³⁺ ions situated at octahedral sites (B sites). In contrast, the inverse spinel structure comprises M³⁺ ions located at A sites and mixed valence M²⁺ and M³⁺ ions located in equal amounts at B sites. Among these oxides, Fe₃O₄ ([Fe³⁺ (S_Fe = 5/2)]_{A site}[Fe³⁺ (S_Fe = 5/2), Fe²⁺ (S_Fe = 2)]_{B site}O₄) – a well-known inverse spinel oxide – was found by Yamada et al.⁷² to exhibit ferrimagnetism (S_total = 2) owing to the antiferromagnetic coupling between the magnetic moments of Fe ions at A sites (S_{Fe, A} = 5/2) and B sites (S_{Fe, B} = 2/5 + 2). Notably, the magnetisation of Fe₃O₄ was controlled by the electrochemical Fe³⁺/Fe²⁺ redox reaction at A and B sites and the structural change induced by lithiation between 2.9 and 1.0 V (Fig. 14(a–c) and (d)). The mechanism underlying the change in magnetisation between 2.9 and 1.3 V was established based on the aforementioned antimagnetic coupling between Fe ions at A and B sites (Fig. 14(e)). At 1.8 V, the Fe³⁺ ions at A sites were reduced to Fe²⁺, leading to an increase in S_total to 2.5 (= −S_{Fe, A} + S_{Fe, B} = −2 + 5/2 + 2). At 1.3 V, 40% of the Fe³⁺ ions at B sites were reduced to Fe²⁺, leading to an increase in S_total to 2.3 (= −S_{Fe, A} + S_{Fe, B} = −2 + 2 × 0.4 + 5/2 × 0.6 + 2). However, at <1.1 V, the magnetisation was irreversibly altered because the structure changed from the inverse spinel structure into the rock-salt configuration Li_1.5Fe₃O₄ and α-Fe nanoparticles (0 V). Other inverse spinel ferrimagnets such as γ-Fe₂O₃,⁷³ MnFe₂O₄,⁷³ Co_0.5Ni_0.5Fe₂O₄,⁷⁴ CoFe₂O₄,⁷⁴ and CuFe₂O₄ ⁷⁵ have also shown similar changes in magnetism. Electrochemically driven ferromagnetism has been realised in lithium-ion batteries using α-Fe₂O₃, which exhibits a non-magnetic corundum crystal structure. α-Fe₂O₃ (Fe³⁺) was converted to LiFeO₂ (Fe³⁺) between 3.0 and 1.0 V, which then transformed into ferromagnetic δ-Fe at voltages below 0.6 V owing to further lithiation. Furthermore, this conversion was reversible, suggesting that electrochemical lithiation–delithiation in batteries could induce ferromagnetism (100 emu g⁻¹, lithiation at 0.5 V) and non-magnetism (10 emu g⁻¹, delithiation at 2.0 V) of α-Fe₂O₃.⁷⁶

Fig. 14 (a–c) Structural changes during electrochemical lithiation of Fe₃O₄.⁷² (d) Temperature dependence of magnetisation for Fe₃O₄ cathodes in a lithium battery equipped with an in situ magnetic measurement system in the voltage range of 1.0–2.9 V.⁷² (e) Spin states during lithiation of Fe₃O₄ at 2.9, 1.8, and 1.3 V.⁷² Reproduced from ref. 72 with permission. Copyright 2014, Royal Society of Chemistry.

Recently, fluoride-ion intercalation has also been performed to alter the magnetism of transition metal oxides. Instead of Li⁺ ions, F⁻ ions can be intercalated into active materials with redox-active, low-valence transition metal ions, such as Mn²⁺,⁷⁷ Fe²⁺,⁷⁸ and Co²⁺;⁷⁹ this has enabled La_2−2xSr_1+2xMn₂O₇ to electrochemically undergo a change in magnetism. La_2−2xSr_1+2xMn₂O₇, which exhibits a Ruddlesden–Popper-type perovskite crystal structure with Mn²⁺ ions (A_n+1B_nO_3n+1), was found to undergo the Mn²⁺/Mn³⁺ redox reaction after fluoride-ion insertion.⁸⁰ The magnetisation of La_2−2xSr_1+2xMn₂O₇ decreased to ∼37% and 6% of the original magnetisation (1st charging to 0.45 V and 0.6 V, respectively, vs. PbF₂/Pb). This difference was caused by the residual fluoride phase, indicating that charging to a higher voltage hindered the recovery of the original magnetisation. Additionally, La_1.3Sr_1.7Mn₂O₇ charged to 0.45 V vs. PbF₂/Pb exhibited 55% of the original magnetisation, although La_1.3Sr_1.7Mn₂O₇ charged to 0.6 V vs. PbF₂/Pb showed only 21% of the original value.

4.3. Changes in magnetism induced by redox reactions of organic ligands in cell-type electrochemical devices

Recently, MOFs comprising metal clusters and redox-active organic linkers have been used as electrochemically switchable magnets. Certain redox-active organic compounds such as 7,7,8,8-tetracyanoquinodimethane (TCNQ), benzoquinone, and tetrathiafulvalene (TTF) inherently generate radicals and electron pairs during a two-step one-electron redox reaction; this feature can be leveraged to switch the magnetic interactions between organic compounds and paramagnetic metal clusters. Taniguchi et al.⁸¹ realised an electrochemically driven change in magnetisation using two types of MOFs with redox-active ligands in lithium-ion batteries equipped with magnetisation measurement systems (Fig. 12(a)). First, a 2D fishnet framework, [{Ru₂(2,3,5,6-F₄PhCO₂)₄}₂(BTDA-TCNQ)]·2(p-xylene), was investigated as an electrochemically driven magnetic MOF. The specimen comprised two paramagnetic {Ru₂(2,3,5,6-F₄PhCO₂)₄} clusters and one BTDA-TCNQ unit, and showed a reversible paramagnetic–ferrimagnetic change based on the two-step one-electron redox reaction mechanism of the BTDA-TCNQ linker (Fig. 17(a) and (b)). Moreover, the porous crystal structure of [{Ru₂(2,3,5,6-F₄PhCO₂)₄}₂(BTDA-TCNQ)]·2(p-xylene) induced the redox reaction of BTDA-TCNQ owing to lithiation into the crystal structure (Fig. 15(c–e)). In the first stage (2.72–2.38 V vs. Li/Li⁺), Li_x[{Ru₂(2,3,5,6-F₄PhCO₂)₄}₂(BTDA-TCNQ)]·2(p-xylene) gradually showed ferrimagnetic behaviour as T_c increased from 42 to 88 K. This magnetic feature was induced by ferrimagnetic interactions between the S = 1 spins of the [Ru^II,II₂] units and the S = 1/2 spins of the BTDA-TCNQ·⁻ segments (Fig. 15(f)). The increase in T_c was attributed to the networks formed by strong exchange interactions in the [–{Ru^II,II₂}–(BTDA-TCNQC)–{Ru^II,II₂}–] networks. In the second step (at 2.26 V vs. Li/Li⁺), the T_c value of Li_x[{Ru₂(2,3,5,6-F₄PhCO₂)₄}₂(BTDA-TCNQ)]·2(p-xylene) decreased to 62 K owing to the formation of diamagnetic BTDA-TCNQ²⁻via two-electron reduction, which induced paramagnetic ordering. After charging to 3.4 V, [{Ru₂(2,3,5,6-F₄PhCO₂)₄}₂(BTDA-TCNQ)]·2(p-xylene) exhibited the same paramagnetic behaviour as that of the original state. [trans-{Ru₂(CF₃CO₂)₂(2,6-(CF₃)₂PhCO₂)₂}₂TCNQ(OMe)₂]·2(p-xylene) (T_c = 62 K)⁸² and [{Ru₂(2,6-F₂PhCO₂)₄}₂TCNQ(OC₂H₄OH)₂] (T_c = 88 K)⁸³ also exhibited increased T_c values based on the aforementioned mechanism during discharge. In particular, an electrochemically reversible change in magnetism for MOFs was first reported for [{Ru^II,II₂(CF₃CO₂)₄}₂(BTDA-TCNQ)]·(p-xylene).⁸⁴ The one-electron redox reaction ferrimagnetism (T_c = 75 K) is observed between 2.9 and 3.4 V. In the case of (TTF)[{Ru^II,II₂(2,3,5,6-F₄PhCO₂)₄}₂(TCNQ)] (T_c = 78 K, 2.3–3.5 V), the electrochemically reversible paramagnetism–ferrimagnetism transition was induced by the interactions between guest TTF˙⁺ ions and TCNQ˙⁻ linkers.⁸⁵ Although the original (TTF)[{Ru^II,II₂(2,3,5,6-F₄PhCO₂)₄}₂(TCNQ)] specimen exhibited paramagnetism owing to the spin Peierls singlet state between guest TTF˙⁺ ions and TCNQ˙⁻ linkers, TTF˙⁺ was reduced to neutral TTF in the discharged state at 2.3 V vs. Li⁺/Li, thereby triggering ferrimagnetism based on the interactions between the TCNQ˙⁻ linkers (S = 1/2) and Ru^II,II₂ clusters (S = 1). This electrochemical magnetisation change was reversible between 2.3 and 3.5 V.

Fig. 15 (a) Representations of [Ru^II,II₂(2,3,5,6-F4PHCO₂)₄] and BTDA-TCNQ with their redox forms BTDA-TCNQC˙− and BTDA-TCNQ²⁻. (b) Spin-ordered states of the neutral (paramagnetic), 1e⁻-filled (ferrimagnetic), and 2e⁻-filled (paramagnetic) forms.⁸¹ (c) Formula unit of [{Ru₂(2,3,5,6-F4PHCO₂)₄}₂(BTDA-TCNQ)]·4CH₂CL₂·2(p-xylene), with solvent molecules and hydrogen atoms omitted for clarity.⁸¹ (d) Packing form projected along the b-axis, with the interstitial p-xylene solvent molecules shown in pale blue. (e) Two-dimensional layer structure in the (10−1) plane.⁸¹ (f) Temperature-dependence of cathode magnetisation for Li_x[{Ru₂(2,3,5,6-F4PHCO₂)₄} (BTDA-TCNQ)]·2(p-xylene) at different equilibrium voltages vs. Li⁺/Li (OCV) (closed circles: field-cooled magnetisation; open circles: remnant magnetisation). Reproduced from ref. 81 with permission. Copyright 2016, Wiley-VCH.

(NBu₄)[MnCr(Cl₂An)₃] also exhibited electrochemically induced magnetisation based on the interactions between a pair of metal ions (M^II and M^III) and 2,5-dichloro-3,6-dihydroxy-1,4-benzoquinonate ligands (chloranilate, Cl₂An²⁻) (Fig. 16(a) and (b)).⁸⁶ Although (NBu₄)[MnCr(Cl₂An)₃] – which comprised Mn²⁺ (S = 5/2), Cr³⁺ (S = 3/2), and redox-active Cl₂An²⁻ (S = 0)/Cl₂An˙³⁻ (S = 1/2) – originally exhibited paramagnetism (T_c = 10 K) owing to the spins of metal ions separated by Cl₂An²⁻, the electrochemical reduction of Cl₂An²⁻ induced ferrimagnetism owing to the magnetic interactions between metal ions and Cl₂An˙³⁻ (Fig. 16(c)). This indicated that the magnetism of (NBu₄)[MnCr(Cl₂An)₃] could be controlled by the spin state of Cl₂An²⁻/Cl₂An˙³⁻. Furthermore, (NBu₄)[MnCr(Cl₂An)₃] exhibited an electrochemically reversible change in magnetisation between 2.35 and 3.1 V vs. Li⁺/Li (Fig. 16(d–f)). Specifically, T_c and the magnetisation variation ΔM varied reversibly during the discharge and charge processes (2.5 and 3.1 V, respectively). Moreover, (NPr₄)₂[Fe₂(Cl₂An)₃] (T_c = 100 K at 1.94 V vs. Li⁺/Li)⁸⁷ and (H₃O)₂(phz)₃[Fe₂(Cl₂An)₃] (T_c = 128 K at 2.67 V vs. Li⁺/Li)⁸⁸ – which comprised Fe³⁺ and Cl₂An²⁻– have also been investigated as electrochemically driven ferrimagnets with high T_c values due to the partial redox reaction of Cl₂An²⁻/Cl₂An˙³⁻ and Fe²⁺/Fe³⁺.

Fig. 16 (a) Redox reaction of CL₂AN²⁻ linkers (left: CL₂AN²⁻ (S = 0); right: CL₂AN˙³⁻ (S = 1/2)).⁸⁶ (b) Schematic of magnetic interactions tuned by electron filling of CL₂AN²⁻ linkers in a lithium battery system. (d) Reversible change of magnetization in the discharge/charge cycles of (NBu₄)[MnCr(CL₂AN)₃] in an Lib system (blue curve: the discharged state, red charged state). (e) Magnetic field dependence of magnetization at 15 K for discharged (2.35 V vs. Li/Li⁺) and charged states (3.1 V vs. Li/Li⁺). The magnetization variation from the initial charged paramagnetic state at 3.1 V vs. Li/Li⁺. (f) Repetitive cycling of discharged and charged Lib voltages (upper), reversible switching of the ferrimagnetic transition temperature (middle), and reversible magnetization variation at 7 T due to phase switching between paramagnetic and ferrimagnetic states (lower). Blue closed circle: the discharged states and red: the charged states.⁸⁶ Reproduced from ref. 86 with permission. Copyright 2017, American Chemical Society.

4.4. The magnetism changes of the perovskite family based on electric field induced ion migration

The perovskite-structured metal oxides SrMO_3−δ and La_xSr_1−xMO₃ (M: transition metal ions; 0 < x < 1, 0 ≤ δ ≤ 0.25) have been reported as electrochemically switchable magnets. The valences of transition metal ions in La_xSr_1−xMO₃ and SrMO_3−δ have been tuned by the concentration of oxygen O²⁻ (δ) or metal ions La³⁺ and Sr²⁺ (x) and other intercalated cations, inducing the magnetism change. First, the redox reaction of M³⁺/M⁴⁺ in SrMO_3−δ based on oxygen insertion/de-insertion induces the change in the magnetism of SrMO_3−δ, accompanied by the structural change between the perovskite SrMO_3−δ and the brownmillerite SrMO_2.5.⁸⁹ Second, the substitution of Sr²⁺ with La³⁺ reduces the valence of the transition metal ions in the magnetism of SrMO₃ from M⁴⁺ to M³⁺ without the structural change, resulting in the magnetism change based on the spin alignment of M⁴⁺ and M³⁺ through oxygen ions.^90,91 In addition, the bilayers composed of SrMO₃ and La_xSr_1−xMO₃ exhibit the interfacial exchange coupling at the SrMO₃/La_xSr_1−xMO₃ interface.⁹² Particularly, the ferromagnetic/antiferromagnetic perovskite bilayer exhibits exchange bias and a shift of the magnetization–magnetic field curve.⁹³ Recently, the perovskite family, SrMO_3−δ monolayers and SrMO_3−δ/La_xSr_1−xMO₃ bilayers, has been investigated as electric-field controllable magnets based on the oxygen insertion/de-insertion mechanism.

4.4.1. Perovskite SrMO_3−δ monolayers. Lu and his co-authors reported the first electric-field controllable magnetic behavior of SrCoO_3−δ using an EDLT.⁹⁴ EDLT inserted proton H⁺ and the oxide ion O²⁻ into brownmillerite SrCoO_2.5 and perovskite SrCoO_3−δ, respectively. This induced a reversible magnetism change among ferromagnetism (SrCoO_3−δ, Co⁴⁺), antiferromagnetism (SrCoO_2.5, Co³⁺), and weak ferromagnetism (HSrCoO_2.5, Co²⁺). More detailed analyses revealed that the crystal structures changed reversibly between SrCoO_3−δ and brownmillerite SrCoO_2.5 with oxygen vacancy channels oriented vertically to the film's surface at a low voltage (Fig. 17(a)–(d)), resulting in the reversible magnetism change (Fig. 17(e) and (f)).⁹⁵ Furthermore, the oxygen vacancy channel ordering depends on the chemical composition of the brownmillerite SrMO_2.5 phase⁹⁶ and voltage,^97,98 which resulted in control over magnetic behaviour.

Fig. 17 (a) Crystal structure of brownmillerite SrCoO_2.5 having horizontal oxygen vacancy channels (OVCs) (H-SCO) (left), perovskite SrCoO_3−δ (P-SCO) (centre), and brownmillerite SrCoO_2.5 having vertical OVCS (V-SCO) (right). High-angle annular dark field-scanning transmission electron microscopy (HAADF-STEM) images of H-SCO (b), P-SCO (c), and V-SCO (d) grown on STO (001) substrates. Magnetization versus field curves for H-SCO (e), P-SCO (f), and V-SCO (g) measured at 100 K.⁹⁵ Reproduced from ref. 95 with permission. Copyright 2022, American Chemical Society.

4.4.2. Perovskite SrMO_3−δ/La_xSr_1−xMO₃ bilayers. The perovskite SrMO_3−δ/La_xSr_1−xMO₃ bilayers have shown electrochemically switchable magnetism different from the perovskite SrMO_3−δ monolayer. For example, a SrFeO_3−δ (weak ferromagnetism)/La_0.7Sr_0.3MnO₃ bilayer exhibited a ferromagnetic-like behavior due to a huge pinning effect on magnetic moments of La_0.7Sr_0.3MnO₃ while SrFeO_3−δ and SrFeO_2.5 showed no significant magnetic signal (Fig. 18(a) and (b)).⁹⁹ In addition, Fig. 18(b) shows that pristine and oxygen-de-inserted phases (SrFeO_2.5/La_0.7Sr_0.3MnO₃ bilayer) exhibited only very tiny exchange bias fields H_EB of 15 and 28 Oe, respectively. On the other hand, the oxygen-inserted phase (SrFeO_3−δ/La_0.7Sr_0.3MnO₃ bilayer) exhibited a prominent shift of the hysteresis loop toward the negative field direction with an H_EB of 150 Oe and more saturation magnetization M_s compared to the pristine and oxygen-de-inserted phases. These results were attributed to the magnetization switching and a strong pinning effect of La_0.7Sr_0.3MnO₃ at the interface. According to Fig. 18(c), the phenomenon was caused by the lateral oxygen diffusion into ordered oxygen vacancy channels in SrFeO_2.5 parallel to the interface, which enables oxygen to migrate fast and results in the prior phase transition at the interface.

Fig. 18 (a) In-plane hysteresis loops of the pristine, negative-, and positive-gated LSMO–SFO_2.5 samples at 10 K. (b) In-plane hysteresis loops of the single-layer SFO_2.5 and SFO_3−δ samples at 10 K. (c) Series of in situ transmission electron microscopy images with respect to time under the negative-biased −2 V gate voltage.⁹⁹ (d) Moment as a function of the field of the P-SCO/LSMO bilayer after being gated with the sequence of initial (0 V), +2 V and −2 V. (e) Moment vs. field curve of the B-SCO/LSMO bilayer in the initial (0 V) state, −2 V gated state, and +2 V gated state.¹⁰⁰ Reproduced from ref. 99 with permission. Copyright 2019, American Chemical Society. Reproduced from ref. 100 with permission. Copyright 2021, American Chemical Society.

Ji and his co-authors investigated the interesting report on the perovskite bilayer using hard ferromagnetic SrCoO_3−δ and soft ferromagnetic La_0.7Sr_0.3MnO₃.¹⁰⁰ The ionic liquid gating induced the reversible phase transition between SrCoO_3−δ (annihilation of oxygen vacancies) and SrCoO_2.5 (generation of oxygen vacancies), which could realize reversible and nonvolatile switching of the exchange spring and exchange bias effect in the SrCoO_3−δ/La_0.7Sr_0.3MnO₃ bilayers (Fig. 18(d) and (e)).

As mentioned above, electrochemical redox reactions of metal ions or ligands have helped elucidate several mechanisms underlying the changes in magnetism and aided in designing magnetic materials. Specifically, MOFs undergo multistep redox reactions involving metal ions, ligands, and guests. These redox levels induce diverse magnetic interactions, enabling the production of electrochemically driven magnets.

5. Electrochromism

5.1 Mechanism governing electrochromism

Electrochromism is a phenomenon in which electrochemical redox reactions cause materials to change colour reversibly. The colours of materials are recognised by neurons sensing light that is reflected from or passes through the materials, and explained based on electron transfer from lower to higher levels. For instance, the colour of polythiophene (Fig. 19(a)) reversibly changes between red (undoped state) and blue (doped state) owing to the appearance and disappearance of the bipolaron band induced by the reversible redox reaction (Fig. 19(b)).¹⁰¹ In the case of methyl viologen,¹⁰² the dication form is red because blue light is absorbed upon electron transition from the HOMO to its upper molecular orbital. After undergoing two-electron reduction, neutral methyl viologen turns blue owing to red light absorption due to electron transfer from the HOMO+1 to its upper level as well as the transformation of the molecular skeleton between the aromatic and quinoid forms. This indicates that the colours of materials can be controlled by tuning their energy gaps. Furthermore, insoluble materials with relatively high molecular weights, such as organic polymers, coordination polymers, and MOFs, are desired for achieving solid-state electrochromic properties. This section explores various solid-state electrochromic materials that operate under the aforementioned mechanisms.

Fig. 19 Relationship between colours of materials and visible light absorption. Change in light absorption based on variation in electron transfer (a) from the valence band–conduction band to the valence band–bipolaron band in polythiophene¹⁰¹ and (b) from the LUMO–HOMO to LUMO+1–HOMO in methyl viologen.¹⁰²

5.2 Solid-state electrochromic materials

5.2.1. Conducting polymers. Conducting polymers have been investigated as major solid-state electrochromic materials and metal-type electronic conductors.¹⁰³ As mentioned in sections 3.1 and 3.2.1, the energy gaps of conducting polymers can be modified by the appearance and disappearance of bipolaron bands after an electrochemical reaction, thereby inducing electrochromism. Various solid-state electrochromic materials have been synthesised since PPy and PT were first used as skeletons. For example, N-substitution of pyrrole can induce a colour variation based on the change in the energy gap. Although PPy and poly(N-alkyl pyrrole)s with an E_g value of 2.7 eV exhibit yellow-to-brown/black electrochromism in their oxidised states,¹⁰⁴ PPy with benzylideneamino groups shows red-to-blue electrochromism in its oxidised state.¹⁰⁵ 3,4-Substitution is effective in allowing polymerisation to proceed only through the 2- and 5-positions and in preventing neutral PT from oxidising in air, resulting in insoluble polymers with fewer structural defects. In fact, poly(3,4-alkylenedioxythiophene) derivatives (Fig. 20(a)) have been developed as stable materials with a lower bandgap (1.6 eV) than those of PPy (2.7 eV) and PT (2.1 eV), and have shown sky-blue-to-transparent electrochromism in their oxidised states (Fig. 20(b)).¹⁰⁶ The low bandgap energy was attributed to oxygen atoms that imparted electron density to PT, leading to a reduction in the bandgap energy owing to the elevation of the HOMO level of the π-conjugated system.¹⁰⁷ Furthermore, the introduction of aromatic molecules other than pyrrole and thiophene into the PPy and PT chains as spacer units is an effective strategy. For instance, the insertion of redox-inactive phenyl or biphenyl groups into the PPy chain lowers the bandgap to 2.3 or 2.4 eV, respectively,¹⁰⁸ indicating that the structural regularities contribute to the lowering of the bandgaps of PPy and poly(p-phenylene) (3.0 eV).

Fig. 20 (a) Molecular structure of poly(3,4-ethylenedioxythiophene). (b) Spectroelectrochemistry of poly(3,4-ethylenedioxythiophene) in different oxidation states, with inset images showing the variation in blue colour at different oxidation levels.¹⁰⁶ Reproduced from ref. 94 with permission. Copyright 2004, Wiley-VCH.

5.2.2. Coordination polymers and MOFs with redox-active ligands or metals. Recently, electrochromic coordination polymers and MOFs have been investigated as materials with electrochemically induced multi-colour attributes. These substances have different crystal structures comprising redox-active ligands and metals, suggesting that multi-colour modulation can be realised by changing the electronic structures of the ligands or metals. In particular, π-conjugated derivatives such as benzoquinone, naphthalenediimide (NDI), and triphenylamine (TPA) have been used as ligands for electrochromic MOFs. The molecular orbital levels can be altered owing to a redox reaction of π-conjugated molecules accompanied by molecular structures, which induces a change in the visible light absorption. Notably, the quinoid form of benzoquinone is transformed into its aromatic counterpart by electrochemical reduction, resulting in a colour change (Fig. 21(a)). Solid-state electrochromism of a benzoquinone derivative has been realised using Ni-MOF 74, which comprises a 2,5-dihydroxyterephthalate ligand and Ni²⁺. In particular, Ni-MOF 74 (Fig. 21(b)) heated at 125 °C on the ITO electrode (Ni-MOF-125) exhibits reversible transmittance modulation between light yellow and reddish brown colours (44.4%) at ∼0 V, whereas Ni-MOF-120 and Ni-MOF-130 show transmittance variations of 14.6% and 5.0% upon heating at 120 and 130 °C; Ni-MOF-125 exhibited the highest active surface area among the specimens and thus showed enhanced electrochromic performance (Fig. 21(c)).¹⁰⁹

Fig. 21 (a) Redox reaction of benzoquinone. (b) Crystal structure of Ni-MOF 74. (c) Coloured and bleached states of Ni-MOF-120, Ni-MOF-125, and Ni-MOF-130.¹⁰⁹ Reproduced from ref. 109 with permission. Copyright 2017, Springer.

NDI derivatives containing redox-active carbonyl groups have also been used as ligands for n-type electrochromic MOFs. The transmittance spectra of neutral and reduced NDI show a distinct transmittance change due to the π–π* transitions of the aromatic units.^110,111 Moreover, Mg-PDI (PDI = perylenetetracarboxylic dianhydride; Fig. 22(a))¹¹² exhibits multistep solid-state electrochromism based on the PDI/PDI˙⁻/PDI²⁻ redox reaction (Fig. 22(b)). The NDI-based Zn MOFs Zn-NDI-74¹¹³ and Zn(NDI-X) (X = H, SEt, and NHEt; SEt = SC₂H₅, and NHEt = NH-C₂H₅)¹¹⁴ and the Zr-NDI MOF¹¹⁵ have also been investigated as electrochromic MOFs based on the redox reactions of NDI derivatives.

Fig. 22 (a) Simulated crystal structure of Mg-PDI. (b) Digital photographs and CIE colour space plots of Mg-PDI thin films, showing the dependence of the colour state on the redox state of the PDI group.¹¹² Reproduced from ref. 112 with permission. Copyright 2022, Cell Press.

TPA derivatives have been utilised as ligands in p-type electrochromic MOFs. In the oxidised state, the cation radical on the nitrogen atom of TPA is stabilised owing to π-delocalisation in the entire molecular skeleton (Fig. 23(a)). The colours of [Zn₂(TTPA)-(SDB)₂·(DMF)(H₂O)]_n (TTPA = tris(4-(1H-1,2,4-triazol-1-yl)phenyl)amine, SDB = 4,4′-sulfonyldibenzoate)¹¹⁶ and Zn₄O(TCA)₂ – which comprises Zn²⁺ and the 4,4′,4′′-tricarboxytriphenylamine (TCA) linker¹¹⁷ – change from transparent to blue at 1.2 V vs. Fc/Fc⁺ and from transparent to dark blue at 1.8 V vs. Ag/AgCl, respectively, owing to the TPA/TPA˙ redox reaction. Furthermore, Zn₄O(TCA)₂ exhibits electrofluorescence because the radical cation serves as a fluorescence quencher, indicating that the radicals on the ligands have the potential to realise electrochemically induced physical properties in a unique manner (Fig. 23(b) and (c)). As mentioned above, immobilisation on MOFs is a powerful method for producing soluble organic materials with solid-state electrochromic properties.

Fig. 23 (a) Redox reaction of TPA. (b) Synthesis route for the Zn₄O(TCA)₂ MOF, and a schematic showing its cathodic electrodeposition.¹¹⁷ (c) Photographs showing coloured and bleached states under bright conditions and fluorescent and nonfluorescent states under dark conditions (excitation under 365 nm) reproduced from ref. 117 with permission. Copyright 2021, American Chemical Society.

Electrochromism has also been observed in coordination polymers and MOFs containing redox-active transition metals. Conventionally, V₂O₅,¹¹⁸ NiO,¹¹⁹ WO₃,¹²⁰ and Nb₂O₅ ¹²¹ have been used as solid-state electrochromic materials with transition metal ions. The electrochromism of these oxides can be tuned by adjusting the doping degree, substitution of metal ions, and oxide-ion content,¹²² which enables them to exhibit multiple colours. In contrast, the solid-state electrochromic properties of coordination polymers such as PB, PBAs, and MOFs can be tuned based on the type of metal ion and the combination of metal ions and linkers. In the case of PB and PBAs, multi-colour electrochromism can be realised through the redox reactions of Fe³⁺/Fe²⁺ (transparent to blue¹²³ and pale green to reddish-brown¹²⁴). Additionally, PBAs with a combination of metal ions such as Cu-PBA (pale yellow to red)¹²⁵ and Ru-PBA (transparent to purple)¹²⁶ exhibit different electrochromic behaviours. Furthermore, MOFs also exhibit different electrochromic effects owing to metal-to-ligand charge transfer (MLCT) based on the combination of metal ions and ligands. A 2D MOF comprising Fe²⁺ ions and 1,3,5-tris(4-(2,2′:6′,2′′-terpyridyl)phenyl)benzene linkers exhibits deep purple-to-pale yellow electrochromism based on the appearance and disappearance of the MLCT band (absorption peak λ_max = 578 nm) owing to the Fe²⁺/Fe³⁺ redox reaction at 1.16 and 0.16 V (Fig. 24(a)–(c)). However, in the case of its counterpart featuring Fe²⁺ ions and electron-withdrawing 1,3,5-tris((2,2′:6′,2′′-terpyridyl)ethynyl)benzene linkers (Fig. 24(a)), the Fe²⁺/Fe³⁺ redox reactions change the colour from deep violet to pale yellow owing to the appearance and disappearance of the MLCT band (λ_max = 588 nm) at 1.16 V (Fig. 24(d) and (e)). Additionally, a 2D MOF comprising Co²⁺ ions and 1,3,5-tris(4-(2,2′:6′,2′′-terpyridyl)phenyl)benzene planar linkers exhibits orange-to-deep purple electrochromism based on the appearance and disappearance of the MLCT band (λ_max = ∼600 nm) owing to the Co²⁺/Co⁺ redox reactions (Fig. 22(f) and (g)).¹²⁷ As mentioned above, coordination polymers and MOFs exhibit outstanding solid-state electrochromic properties owing to their insolubility and redox-active components. Additionally, diverse MOF structures can be viably used to construct chromic devices.¹²⁷

Fig. 24 (a) Structures of bis(terpyridine)metal(II) complex nanosheets, 1,3,5-tris(4-(2,2′:6′,2′′-terpyridyl)phenyl)benzene, and 1,3,5-tris((2,2′:6′,2′′-terpyridyl)ethynyl)benzene ligands.¹¹⁵ (b, d and f) Electrochromic behaviour and (c, e and g) UV–vis spectra of (b and c) nanosheets comprising Fe²⁺ and 1,3,5-tris(4-(2,2′:6′,2′′-terpyridyl)phenyl)benzene on transparent ITO electrodes, (d and e) nanosheets composed of Fe²⁺ and 1,3,5-tris((2,2′:6′,2′′-terpyridyl)ethynyl)benzene on transparent ITO electrodes, and (f and g) nanosheets comprising Co²⁺ and 1,3,5-tris(4-(2,2′:6′,2′′-terpyridyl)phenyl)benzene on transparent ITO electrodes.¹²⁷ Reproduced from ref. 127 with permission. Copyright 2015, American Chemical Society.

Overall, conducting polymers and organic–inorganic hybrid materials have been investigated as excellent electrochromic materials. These materials have diverse molecular skeletons and crystal structures with organic compounds and metal ions, resulting in various types of electrochromic behaviours that can potentially be applied in feasible electrochromic devices.

6. Conclusions

The physical properties of transition metal compounds such as conductivity, magnetism, and colour can be controlled by electrochemical redox reactions. Changes in these physical properties can be induced by tweaking the electronic state of the transition metal or its surroundings. Additionally, interactions based on molecular or crystal structures are important for the design of external-stimuli-responsive materials. The insights provided herein are anticipated to aid in controlling functional materials such as ferroelectrics using electrochemical devices and developing new physical properties such as electrofluoromism¹²⁸ and putting them into practical use.

Outlook

The previous findings contribute to both the discovery and application of solid-state materials with electrochemically switchable physical properties. Recently, chiroptical properties of molecules have been switched in solution by electrochemical reactions.¹²⁹ Furthermore, redox-switchable ion flow¹³⁰ and wettability¹³¹ also have been reported. We believe that the solid-state materials, such as polymers and MOFs, can realize solid-state electrochemically switchable chiroptical properties and that the phenomena can be applied to functional devices, such as ion diodes, controlled-release plates, and so on.

Data availability

No primary research results, software or code have been included and no new data were generated or analysed as part of this review.

Author contributions

Conceptualization: Takeshi Shimizu and Hirofumi Yoshikawa; investigation: Takeshi Shimizu and Shunsuke Ohkata; supervision: Naoki Tanifuji and Hirofumi Yoshikawa; writing – original draft: Takeshi Shimizu; writing – review & editing: Katsuhiro Wakamatsu, Wang Heng, Naoki Tanifuji, and Hirofumi Yoshikawa.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This study was supported by a Japan Society for the Promotion of Science (JSPS) KAKENHI Grant-in-Aid for Early-Career Scientists (no. 23K13830), JSPS KAKENHI Grant-in-Aid for Scientific Research (B) (no. 23H02041 and 23K26734), a JSPS KAKENHI Grant-in-Aid for Challenging Research (Exploratory) (no. 22K19073), and the GEAR 5.0 Project of the National Institute of Technology (KOSEN) in Japan. The authors also acknowledge financial support from The Mazda Foundation (23KK-146), the Futaba Foundation, Yazaki Memorial Foundation for Science and Technology, and the JSPS Core-to-Core Program A—Advanced Research Networks “International Network on Polyoxometalate Science for Advanced Functional Energy Materials”, Collaborative Special Research Subsidy and Individual Research Subsidy (Kwansei Gakuin University, 2021).

References

1. M. Mrinalini and S. Prasanthkumar, ChemPlusChem, 2019, 84, 1103–1121.
2. W. Cai, J. Wang, C. Chu, W. Chen, C. Wu and G. Liu, Adv. Sci., 2019, 6, 1801526.
3. Z. M. Png, C. G. Wang, J. C. C. Yeo, J. J. C. Lee, N. E. Surat'man, Y. L. Tan, H. Liu, P. Wang, B. H. Tan, J. W. Xu, X. J. Loh and Q. Zhu, Mol. Syst. Des. Eng., 2023, 8, 1097–1129.
4. T. Bukreeva, V. Barachevsky, O. Venidiktova, P. Krikunova and T. Pallaeva, Mater. Today Commun., 2024, 38, 107769.
5. Y. Z. Huang, R. K. Gupta, G. G. Luo, Q. C. Zhang and D. Sun, Coord. Chem. Rev., 2024, 499, 215508.
6. F. Zhang, X. Miao, N. van Dijk, E. Brück and Y. Ren, Adv. Energy Mater., 2024, 2400369,  DOI:10.1002/aenm.202400369.
7. B. Xu, J. Chen, Z. Ding, J. Hu, Y. Zhang, H. Li and H. Wang, Small Sci., 2023, 3, 2300025.
8. W. M. Huang, Y. Zhao, C. C. Wang, Z. Ding, H. Purnawali, C. Tang and J. L. Zhang, J. Polym. Res., 2012, 19, 9952.
9. Y. Liu, C. Xiao, P. Huang, M. Cheng and Y. Xie, Chem, 2018, 4, 1263–1283.
10. S. A. Lee, J. W. Yang, S. Choi and H. W. Jang, Exploration, 2021, 1, 20210012.
11. S. Lany, J. Phys.: Condens. Matter, 2015, 27, 283203.
12. M. D. Allendorf, C. A. Bauer, R. K. Bhakta and R. J. T. Houk, Chem. Soc. Rev., 2009, 38, 1330–1352.
13. T. Gorai, W. Schmitt and T. Gunnlaugsson, Dalton Trans., 2021, 50, 770–784.
14. C. Leighton, Nat. Mater., 2019, 18, 13–18.
15. S. R. Forrest, Nature, 2004, 428, 911–918.
16. H. Liu, Y. Wang, Z. Qin, D. Liu, H. Xu, H. Dong and W. Hu, J. Phys. Chem. Lett., 2021, 12, 1612–1630.
17. M. Heydari Gharahcheshmeh and K. K. Gleason, Mater. Today Adv., 2020, 8, 100086.
18. C. Li, X. Sun, Y. Yao and G. Hong, Mater. Today Nano, 2021, 13, 100105.
19. H. Shirakawa, T. Ito and S. Ikeda, Makromol. Chem., 1978, 179, 1565–1573.
20. P. J. Nigrey, A. G. MacDiarmid and A. J. Heeger, J. Chem. Soc., Chem. Commun., 1979, 594–595.
21. G. C. Farrington, B. Scrosati, D. Frydrych and J. DeNuzzio, J. Electrochem. Soc., 1984, 131, 7–12.
22. N. Noma, Y. Shirota and H. Mikawa, Nippon Kagaku Kaishi, 1986, 3, 312–318.
23. I. R. Jeon, N. Noma, R. F. C. Claridge and Y. Shirota, Polym. J., 1992, 24, 273–279.
24. A. A. Iurchenkova, T. Kallio and E. O. Fedorovskaya, Electrochim. Acta, 2021, 391, 138892.
25. T. C. Chung, J. H. Kaufman, A. J. Heeger and F. Wudl, Phys. Rev. B: Condens. Matter Mater. Phys., 1984, 30, 702–710.
26. P. M. McManus, S. C. Yang and R. J. Cushman, J. Chem. Soc., Chem. Commun., 1985, 1556–1557.
27. B. Zayat, P. Das, B. C. Thompson and S. R. Narayan, J. Phys. Chem. C, 2021, 125, 7533–7541.
28. S. N. Patel, A. E. Javier and N. P. Balsara, ACS Nano, 2013, 7, 6056–6068.
29. F. Bigdeli, C. T. Lollar, A. Morsali and H. C. Zhou, Angew. Chem., Int. Ed., 2020, 59, 4652–4669.
30. L. S. Xie, G. Skorupskii and M. Dincǎ, Chem. Rev., 2020, 120, 8536–8580.
31. X. Sun, K. H. Wu, R. Sakamoto, T. Kusamoto, H. Maeda, X. Ni, W. Jiang, F. Liu, S. Sasaki, H. Masunaga and H. Nishihara, Chem. Sci., 2017, 8, 8078–8085.
32. L. Liu, J. A. DeGayner, L. Sun, D. Z. Zee and T. D. Harris, Chem. Sci., 2019, 10, 4652–4661.
33. H. C. Wentz, G. Skorupskii, A. B. Bonfim, J. L. Mancuso, C. H. Hendon, E. H. Oriel, G. T. Sazama and M. G. Campbell, Chem. Sci., 2020, 11, 1342–1346.
34. Y. Kobayashi, B. Jacobs, M. D. Allendorf and J. R. Long, Chem. Mater., 2010, 22, 4120–4122.
35. I. R. Jeon, L. Sun, B. Negru, R. P. Van Duyne, M. Dinca and T. D. Harris, J. Am. Chem. Soc., 2016, 138, 6583–6590.
36. M. L. Aubrey, B. M. Wiers, S. C. Andrews, T. Sakurai, S. E. Reyes-Lillo, S. M. Hamed, C. J. Yu, L. E. Darago, J. A. Mason, J. O. Baeg, F. Grandjean, G. J. Long, S. Seki, J. B. Neaton, P. Yang and J. R. Long, Nat. Mater., 2018, 17, 625–632.
37. N. Ogihara, N. Ohba and Y. Kishida, Sci. Adv., 2017, 3, e1603103.
38. N. Ogihara, T. Yasuda, Y. Kishida, T. Ohsuna, K. Miyamoto and N. Ohba, Angew. Chem., Int. Ed., 2014, 53, 11467–11472.
39. S. Goswami, I. Hod, J. D. Duan, C. W. Kung, M. Rimoldi, C. D. Malliakas, R. H. Palmer, O. K. Farha and J. T. Hupp, J. Am. Chem. Soc., 2019, 141, 17696–17702.
40. W. Morris, B. Volosskiy, S. Demir, F. Gándara, P. L. McGrier, H. Furukawa, D. Cascio, J. F. Stoddart and O. M. Yaghi, Inorg. Chem., 2012, 51, 6443–6445.
41. P. A. Herrera-Herrera, E. Rodríguez-Sevilla and A. S. Varela, Dalton Trans., 2021, 50, 16939–16944.
42. J. Duan, S. Goswami, S. Patwardhan and J. T. Hupp, J. Phys. Chem. C, 2022, 126, 4601–4611.
43. X. Gui, B. Lv and W. Xie, Chem. Rev., 2021, 121, 2966–2991.
44. J. G. Bednorz and K. A. Muller, Rev. Mod. Phys., 1988, 60, 585–600.
45. W. I. F. David, W. T. A. Harrison, J. M. F. Gunn, O. Moze, A. K. Soper, P. Day, J. D. Jorgensen, D. G. Hinks, M. A. Beno, L. Soderholm, D. W. Capone II, I. K. Schuller, C. U. Segre, K. Zhang and J. D. Grace, Nature, 1987, 327, 310–312.
46. A. H. Salama, A. M. Youssef, Y. S. Rammah and M. El-Khatib, Bull. Natl. Res. Cent., 2019, 43, 89.
47. K. Maiti, J. Fink, S. De Jong, M. Gorgoi, C. Lin, M. Raichle, V. Hinkov, M. Lambacher, A. Erb and M. S. Golden, Phys. Rev. B: Condens. Matter Mater. Phys., 2009, 80, 1–16.
48. T. Oda, M. Shirai, N. Suzuki and K. Motizuki, Physica B: Condens. Matter, 1996, 219–220, 451–453.
49. D. C. Johnston, H. Prakash, W. H. Zachariasen and R. Viswanathan, Mater. Res. Bull., 1973, 8, 777–784.
50. S. Satpathy and R. M. Martin, Phys. Rev. B: Condens. Matter Mater. Phys., 1987, 36, 7269–7272.
51. J. Wang, L. Liu, W. Zhang, F. Dang, S. Zhang and Y. Du, Physica B: Condens. Matter, 2022, 639, 413959.
52. J. Zhao, Q. Shi, Y. Xiang and Y. Xia, Int. J. Electrochem. Sci., 2018, 13, 1921–1930.
53. S. Hamada, M. Kato, T. Noji and Y. Koike, Physica C, 2010, 470, S766–S767.
54. K. Yoshimatsu, M. Niwa, H. Mashiko, T. Oshima and A. Ohtomo, Sci. Rep., 2015, 5, 16325.
55. S. Maruyama, J. Shin, X. Zhang, R. Suchoski, S. Yasui, K. Jin, R. L. Greene and I. Takeuchi, Appl. Phys. Lett., 2015, 107, 142602.
56. Z. Wei, Q. Li, B. C. Gong, X. Wei, W. Hu, Z. Ni, G. He, M. Qin, A. Kusmartseva, F. V. Kusmartsev, J. Yuan, B. Zhu, Q. Chen, J. H. Chen, K. Liu and K. Jin, Phys. Rev. B, 2021, 103, L140501.
57. M. Kato, T. Kajita, R. Hanakago and Y. Koike, Phys. C, 2006, 445–448, 26–30.
58. A. Palau, A. Fernandez-Rodriguez, J. C. Gonzalez-Rosillo, X. Granados, M. Coll, B. Bozzo, R. Ortega-Hernandez, J. Suñé, N. Mestres, X. Obradors and T. Puig, ACS Appl. Mater. Interfaces, 2018, 10, 30522–30531.
59. K. Ueno, S. Nakamura, H. Shimotani, H. T. Yuan, N. Kimura, T. Nojima, H. Aoki, Y. Iwasa and M. Kawasaki, Nat. Nanotechnol., 2011, 6, 408–412.
60. K. Ueno, S. Nakamura, H. Shimotani, A. Ohtomo, N. Kimura, T. Nojima, H. Aoki, Y. Iwasa and M. Kawasaki, Nat. Mater., 2008, 7, 855–858.
61. J. S. Miller, Chem. Soc. Rev., 2011, 40, 3266–3296.
62. V. Sivakumar, C. A. Ross, N. Yabuuchi, Y. Shao-Horn, K. Persson and G. Ceder, J. Electrochem. Soc., 2008, 155, P83.
63. V. Srivastava, M. Rajagopalan and S. P. Sanyal, Eur. Phys. J. B, 2008, 61, 131–139.
64. D. W. Murphy, F. J. Di Salvo, J. N. Carides and J. V. Waszczak, Mater. Res. Bull., 1978, 13, 1395–1402.
65. T. Gamze Ulusoy Ghobadi, E. Ozbay and F. Karadas, Chem. – Eur. J., 2021, 27, 3638–3649.
66. O. Sato, S. Hayami, Y. Einaga and Z. Z. Gu, Bull. Chem. Soc. Jpn., 2003, 76, 443–470.
67. A. Fujishima and K. Hashimoto, Science, 1996, 271, 5–7.
68. T. Yamada, K. Morita, H. Wang, K. Kume, H. Yoshikawa and K. Awaga, Angew. Chem., Int. Ed., 2013, 52, 6238–6241.
69. O. Sato, J. Solid State Electrochem., 2007, 11, 773–779.
70. M. Okubo, D. Asakura, Y. Mizuno, T. Kudo, H. Zhou, A. Okazawa, N. Kojima, K. Ikedo, T. Mizokawa and I. Honma, Angew. Chem., Int. Ed., 2011, 50, 6269–6273.
71. C. H. Li, M. K. Peprah, D. Asakura, M. W. Meisel, M. Okubo and D. R. Talham, Chem. Mater., 2015, 27, 1524–1530.
72. T. Yamada, K. Morita, K. Kume, H. Yoshikawa and K. Awaga, J. Mater. Chem. C, 2014, 2, 5183–5188.
73. G. Wei, L. Wei, D. Wang, Y. Chen, Y. Tian, S. Yan, L. Mei and J. Jiao, Sci. Rep., 2017, 7, 12554.
74. L. A. Dubraja, C. Reitz, L. Velasco, R. Witte, R. Kruk, H. Hahn and T. Brezesinski, ACS Appl. Nano Mater., 2018, 1, 65–72.
75. S. Dasgupta, B. Das, Q. Li, D. Wang, T. T. Baby, S. Indris, M. Knapp, H. Ehrenberg, K. Fink, R. Kruk and H. Hahn, Adv. Funct. Mater., 2016, 26, 7507–7515.
76. Q. Zhang, X. Luo, L. Wang, L. Zhang, B. Khalid, J. Gong and H. Wu, Nano Lett., 2016, 16, 583–587.
77. M. A. Nowroozi, K. Wissel, J. Rohrer, A. R. Munnangi and O. Clemens, Chem. Mater., 2017, 29, 3441–3453.
78. M. A. Nowroozi, B. de Laune and O. Clemens, ChemistryOpen, 2018, 7, 617–623.
79. M. A. Nowroozi, S. Ivlev, J. Rohrer and O. Clemens, J. Mater. Chem. A, 2018, 6, 4658–4669.
80. S. Vasala, A. Jakob, K. Wissel, A. I. Waidha, L. Alff and O. Clemens, Adv. Electron. Mater., 2020, 6, 1900974.
81. K. Taniguchi, K. Narushima, J. Mahin, W. Kosaka and H. Miyasaka, Angew. Chem., Int. Ed., 2016, 55, 5238–5242.
82. K. Taniguchi, N. Shito, H. Fukunaga and H. Miyasaka, Chem. Lett., 2018, 47, 664–667.
83. K. Taniguchi, K. Narushima, K. Yamagishi, N. Shito, W. Kosaka and H. Miyasaka, Jpn. J. Appl. Phys., 2017, 56, 060307.
84. K. Taniguchi, K. Narushima, H. Sagayama, W. Kosaka, N. Shito and H. Miyasaka, Adv. Funct. Mater., 2017, 27, 1604990.
85. H. Fukunaga, M. Tonouchi, K. Taniguchi, W. Kosaka, S. Kimura and H. Miyasaka, Chem. – Eur. J., 2018, 24, 4294–4303.
86. K. Taniguchi, J. Chen, Y. Sekine and H. Miyasaka, Chem. Mater., 2017, 29, 10053–10059.
87. J. Chen, K. Taniguchi, Y. Sekine and H. Miyasaka, Inorg. Chem., 2021, 60, 9456–9460.
88. J. Chen, K. Taniguchi, Y. Sekine and H. Miyasaka, J. Magn. Magn. Mater., 2020, 494, 165818.
89. H. Jeen, W. S. Choi, M. D. Biegalski, C. M. Folkman, I. C. Tung, D. D. Fong, J. W. Freeland, D. Shin, H. Ohta, M. F. Chisholm and H. N. Lee, Nat. Mater., 2013, 12, 1057–1063.
90. M. Kawasaki and Y. Tokura, Phys. Rev. B: Condens. Matter Mater. Phys., 2003, 61, 12187–12195.
91. C. Zener, Phys. Rev., 1951, 81, 403–405.
92. J. F. Ding, O. I. Lebedev, S. Turner, Y. F. Tian, W. J. Hu, J. W. Seo, C. Panagopoulos, W. Prellier, G. Van Tendeloo and T. Wu, Phys. Rev. B: Condens. Matter Mater. Phys., 2013, 87, 1–7.
93. J. Zhang, G. Zhou, Z. Yan, H. Ji, X. Li, Z. Quan, Y. Bai and X. Xu, ACS Appl. Mater. Interfaces, 2019, 11, 26460–26466.
94. N. Lu, P. Zhang, Q. Zhang, R. Qiao, Q. He, H. B. Li, Y. Wang, J. Guo, D. Zhang, Z. Duan, Z. Li, M. Wang, S. Yang, M. Yan, E. Arenholz, S. Zhou, W. Yang, L. Gu, C. W. Nan, J. Wu, Y. Tokura and P. Yu, Nature, 2017, 546, 124–128.
95. H. Han, A. Sharma, H. L. Meyerheim, J. Yoon, H. Deniz, K. R. Jeon, A. K. Sharma, K. Mohseni, C. Guillemard, M. Valvidares, P. Gargiani and S. S. P. Parkin, ACS Nano, 2022, 16, 6206–6214.
96. A. Khare, J. Lee, J. Park, G. Y. Kim, S. Y. Choi, T. Katase, S. Roh, T. S. Yoo, J. Hwang, H. Ohta, J. Son and W. S. Choi, ACS Appl. Mater. Interfaces, 2018, 10, 4831–4837.
97. V. Chaturvedi, W. M. Postiglione, R. D. Chakraborty, B. Yu, W. Tabiś, S. Hameed, N. Biniskos, A. Jacobson, Z. Zhang, H. Zhou, M. Greven, V. E. Ferry and C. Leighton, ACS Appl. Mater. Interfaces, 2021, 13, 51205–51217.
98. B. Cui, Y. Huan and J. Hu, J. Phys.: Condens. Matter, 2020, 32, 344001.
99. M. S. Saleem, B. Cui, C. Song, Y. Sun, Y. Gu, R. Zhang, M. U. Fayaz, X. Zhou, P. Werner, S. S. P. Parkin and F. Pan, ACS Appl. Mater. Interfaces, 2019, 11, 6581–6588.
100. H. Ji, G. Zhou, X. Wang, J. Zhang, P. Kang and X. Xu, ACS Appl. Mater. Interfaces, 2021, 13, 15774–15782.
101. R. J. Mortimer, D. R. Rosseinsky and P. M. S. Monk, Electrochromic Materials and Devices, Wiley-VCH, Hoboken, New Jersey, 2015.
102. G. Sonmez, Chem. Commun., 2005, 5251–5259.
103. T. V. Nguyen, Q. V. Le, S. Peng, Z. Dai, S. H. Ahn and S. Y. Kim, Adv. Mater. Technol., 2023, 8, 2300474.
104. A. F. Diaz, J. Castillo, K. K. Kanazawa, J. A. Logan, M. Salmon and O. Fajardo, J. Electroanal. Chem. Interfacial Electrochem., 1982, 133, 233–239.
105. Y. Murakami and T. Yamamoto, Polym. J., 1999, 31, 476–478.
106. G. Sonmez, H. B. Sonmez, C. K. F. Shen and F. Wudl, Adv. Mater., 2004, 16, 1905–1908.
107. D. M. Welsh, A. Kumar, E. W. Meijer and J. R. Reynolds, Adv. Mater., 1999, 11, 1379–1382.
108. G. A. Sotzing, J. R. Reynolds, A. R. Katritzky, J. Soloducho, S. Belyakov and R. Musgrave, Macromolecules, 1996, 29, 1679–1684.
109. N. Zhang, Y. Jin, Q. Zhang, J. Liu, Y. Zhang and H. Wang, Ionics, 2021, 27, 3655–3662.
110. S. Halder, S. Roy and C. Chakraborty, Sol. Energy Mater. Sol. Cells, 2022, 234, 111429.
111. S. H. Hsiao and Y. Z. Chen, J. Electroanal. Chem., 2017, 799, 417–423.
112. Z. Lu, R. Li, L. Ping, Z. Bai, K. Li, Q. Zhang, C. Hou, Y. Li, W. Jin, X. Ling and H. Wang, Cell Rep. Phys. Sci., 2022, 3, 100866.
113. X. Wu, K. Wang, J. Lin, D. Yan, Z. Guo and H. Zhan, J. Colloid Interface Sci., 2021, 594, 73–79.
114. C. R. Wade, M. Li and M. Dincǎ, Angew. Chem., Int. Ed., 2013, 52, 13377–13381.
115. G. Radha, S. Roy, C. Chakraborty and H. Aggarwal, Chem. Commun., 2022, 58, 4024–4027.
116. C. M. Ngue, Y. H. Liu, M. K. Leung and K. L. Lu, Inorg. Chem., 2021, 60, 11458–11465.
117. J. Liu, X. Y. D. Ma, Z. Wang, L. Xu, F. Wang, C. He and X. Lu, ACS Appl. Electron. Mater., 2021, 3, 1489–1495.
118. I. Mjejri, M. Gaudon and A. Rougier, Sol. Energy Mater. Sol. Cells, 2019, 198, 19–25.
119. G. A. Niklasson and C. G. Granqvist, J. Mater. Chem., 2007, 17, 127–156.
120. S. H. Baeck, K. S. Choi, T. F. Jaramillo, G. D. Stucky and E. W. McFarland, Adv. Mater., 2003, 15, 1269–1273.
121. R. Romero, E. A. Dalchiele, F. Martín, D. Leinen and J. R. Ramos-Barrado, Sol. Energy Mater. Sol. Cells, 2009, 93, 222–229.
122. V. Rai, R. S. Singh, D. J. Blackwood and D. Zhili, Adv. Eng. Mater., 2020, 22, 2000082.
123. K. Itaya, H. Akahoshi and S. Toshima, J. Electrochem. Soc., 1982, 129, 1498–1500.
124. C. F. Yang, Q. Wang, C. Y. Yi, J. H. Zhao, J. Fang and W. G. Shen, J. Electroanal. Chem., 2012, 674, 30–37.
125. M. K. Sharma and S. K. Aggarwal, J. Electroanal. Chem., 2013, 705, 64–67.
126. R. J. Mortimer and T. S. Varley, Sol. Energy Mater. Sol. Cells, 2013, 109, 275–279.
127. K. Takada, R. Sakamoto, S. T. Yi, S. Katagiri, T. Kambe and H. Nishihara, J. Am. Chem. Soc., 2015, 137, 4681–4689.
128. P. Audebert and F. Miomandre, Chem. Sci., 2013, 4, 575–584.
129. J. W. Canary, Chem. Soc. Rev., 2009, 38, 747–756.
130. Q. Zhang, Z. Zhang, H. Zhou, Z. Xie, L. Wen, Z. Liu, J. Zhai and X. Diao, Nano Res., 2017, 10, 3715–3725.
131. J. Lahann, S. Mitragotri, T. N. Tran, H. Kaido, J. Sundaram, I. S. Choi, S. Hoffer, G. A. Somorjai and R. Langer, Science, 2003, 299, 371–374.

---