Photoswitchable high-dimensional Co^II–[W^V(CN)₈] networks: Past, present, and future

Photoswitching of magnetic properties is a subject of modern interest in materials science.^1–19 A successful approach is to tune the electronic state of magnetic materials. For example, magnetization is controlled when the oxidation numbers of transition metal ions within a magnetic material are varied by photoirradiation. In addition, the bistability of the electronic states is indispensable for observing photoinduced magnetization since the photogenerated state must endure even after photoirradiation is terminated. Along these lines, the photomagnetic properties of cyanido-bridged metal assemblies have been studied.^20–36 For example, the photodemagnetization in iron(II)-hexacyanidochromate(III)^37–40 and rubidium-manganese(II)-hexacyanidoferrate(III),^41–47 the photoinduced magnetization in cobalt(II)-hexacyanidoferrate(III),^6,48–61 manganese(II)-octacyanidotungstate(IV),^62,63 iron(II)-octacyanidoniobate(IV),^64–66 cobalt(II)-octacyanidotungstate(V),^67–74 and copper(II)-octacyanidomolybdate(IV),^75–113 and a photoinduced magnetic pole inversion have been achieved in iron(II)-manganese(II)-hexacyanidochromate(III). Furthermore, cyanido-bridged metal assemblies have been studied as functionalized molecule-based magnets. Octacyanidometallates [M(CN)₈]ⁿ⁻ (M = Mo, Nb, Re, or W) are useful building blocks for functionalized molecule-based magnets since they can adopt three different spatial configurations [e.g., square antiprism (D_4h), dodecahedron (D_2d), and bicapped trigonal prism (C_2v)], depending on their chemical environment, such as the surrounding ligands.^114–118 In particular, the photomagnetism of M^II–[M^IV/V(CN)₈] systems has been attracting attention recently. The spin crossover in the M^II (i.e., Fe) as well as M^IV/V (i.e., Mo and W) sites, and electron charge transfer between 3d transition metals and octacyanidometallates, bring the variety in photoinduced phase transitions.

Research on the mechanism of thermal-induced spin transition in Prussian blue analogues (PBAs) is equally essential. Formerly, this mechanism was instinctively attributed to a charge-transfer-induced spin transition (CTIST), e.g., transition from high temperature phase Co^II–HSN=C—W^V to low temperature one Co^III–LS—N=C—W^IV for L—Co—[W(CN)₈] systems.⁶⁷ Nevertheless, the most recent results of femtosecond x-ray and optical absorption spectroscopies for A_xCo_y[Fe(CN)₆]⋅nH₂O, A = alkali cation, confirmed that thermal-induced spin transition is initiated by the spin transition on cobalt center (Co^II–HS to Co^II–LS) and the related structural reorganizations induces the charge transfer in cobalt–iron pair (Co^II–LS–Fe^III–LS to Co^III–LS–Fe^II–LS).⁶¹ Bearing in mind the insufficient amount of data confirming the correctness of the CTIST mechanism for Co^II-[W^V(CN)₈] systems, in this Perspective, we will use the phrase: electron-transfer-coupled spin transition, and the answer to the question which mechanism is correct is left as one of the future challenges.

In this Perspective, we are going to introduce the most of a vivid instance of high-dimensional photomagnetic systems-based L—Co^II—[W^V(CN)₈] networks, L = organic ligand, revealing the electron-transfer-coupled spin transition associated with a thermally reversible conversion between paramagnetic state Co^II–HS(S = ³/₂)—N=C—W^V(S = ¹/₂) at high temperature and diamagnetic one Co^III–LS(S = 0)—N=C—W^IV(S = 0) at low temperature and photoinduced spontaneous magnetization with large thermal hysteresis.^67–74 

Historically, the first report of functional high-dimensional Co^II–[W^V(CN)₈] coordination polymer can be set in 2003.⁶⁷ At that time, a new type of Prussian blue analogues, Cs^I{[Co^II(3-CNpy)₂][W^V(CN)₈]} ⋅ H₂O where 3-CNpy = 3-cyanopyridine, has been reported by Ohkoshi et al. This two-dimensional (2D) assembly consists of cyanido-bridged folded square grids with four-metallic units separated by charge compensating Cs⁺ cations and neutral water molecules [Fig. 1(a)]. Each monolayer is built of octacyanidotungstate(V) anions linked by four cyanide to four cobalt(II) centers. Meanwhile, each Co(II) center is equatorial cyanido-bridged to four anions and axial coordinated by two nitrogen atoms of 3-cyanopyridine molecules. The average interlayers distance adopts a value of 12.42 Å.

Magnetic studies of Cs^I{[Co^II(3-CNpy)₂][W^V(CN)₈]} ⋅ H₂O showed that this compound exhibits the electron-transfer-coupled spin transition with a 49 K wide thermal hysteresis loop. The transition from high-temperature (HT) to low-temperature (LT) phase and vice versa occur at 167 and 216 K, respectively [Fig. 2(a)]. This observation has been also confirmed by temperature-dependent IR spectroscopy, and it indicated that the discovered temperature-induced phase transition from HT to LT phases is due to the electronic state change from Co^II–HS(S = ³/₂)—N=C—W^V(S = ¹/₂) to Co^III–LS(S = 0)—N=C—W^IV(S = 0). The quintessence of this work was detection of the photomagnetic effect induced by red light (600–750 nm), corresponding to intervalence-transfer (IT) band between Co^III–W^IV and Co^II–W^V. After excitation for 4 h at 5 K, the sample revealed drastic increase in magnetization correlated with the occurrence of long-range ferromagnetic ordering with Curie temperature (T_C) of 30 K [Fig. 2(b)]. Thermal treating up to 120 K leads to recovery of the compound to the initial state; thus, this system can work in repetitive cycles of irradiation and heating. Additionally, the IR and powder x-ray diffraction studies of the photoexcited phase at 10 K confirmed that the photomagnetic effect is associated with a charge transfer from the LT cobalt(III)-octacyanidotungstate(IV) phase to the HT cobalt(II)-octacyanidotungstate(V) one, and it is caused by bistability due to the changes of the valence state and coordination geometry of [W^V(CN)₈]³⁻ anion.

A few years later, two similar three-dimensional (3D) assemblies with general formula of [Co^II₃(pym)₄[W^V(CN)₈]₂ ⋅ 6H₂O, pym = pyrimidine, have been developed.^68–70 The structural unit of the first form (I) consists of octacyanidotungstate(V) anions bridged to four [Co1(pym)(H₂O)]²⁺ and one [Co2(H₂O)₂]²⁺ complex cations [Fig. 3(a)]. The four equatorial positions of Co1 are occupied by cyanide nitrogen atoms, while the apical positions are coordinated by the pyrimidine nitrogen atom and the water oxygen atom. The Co2 center is surrounded by two cyanide nitrogen atoms, two pyrimidine nitrogen atoms, and two water oxygen atoms. Moreover, the cyanido-bridged Co1-W1 folded square grids with four-metallic units along the ab-plane are stabilized by Co1-pym-Co2 and CN-Co2-NC bridges and crystallization pyrimidine molecules filling the voids in layers. The structure of the second form (II) differs only in that the coordination water of [Co1(pym)(H₂O)]²⁺ entity has been swapped with the crystallization pyrimidine molecule, generating new local environment of [Co1(pym)₂]²⁺ without the impact on the number of cyano-bridges and topology of structure [Fig. 3(b)]. The closest interlayer distances adopt values of 11.08 and 11.34 Å for forms I and II, respectively.

This simple exchange of crystallization ligands has a strong impact on magnetic properties of both forms. Compound I exhibits long-range ferromagnetic ordering with a Curie temperature of 32 K and magnetic hysteresis loop with the coercive field (H_C) of 12 kOe at 2 K. Interestingly, the magnetic studies for II material show similar electron-transfer-coupled spin transition to the previous Cs^I{[Co^II(3-CNpy)₂][W^V(CN)₈]} ⋅ H₂O system, however, with almost twice broader hysteresis of 90 K (T_1/2^{HT to LT} = 208 K and T_1/2^{LT to HT} = 298 K) [Fig. 4(a)]. This effect has been also verified by temperature-dependent IR spectroscopy. Moreover, magnetic properties have been assigned to the electron-transfer-coupled spin transition from HT phase Co^II–HS(S = ³/₂)—N=C—W^V(S = ¹/₂) to LT phase Co^III–LS(S = 0)—N=C—W^IV(S = 0). Furthermore, this compound exhibits the photomagnetic phenomenon induced by 840 nm continuous-wave diode laser irradiation for 160 min at 5 K. After illumination with 840 nm light, the sample exhibits a spontaneous magnetization with a T_C of 40 K [Fig. 4(b)] and enormously large magnetic hysteresis loop for photomagnetic coordination compounds with the coercive field (H_C) of 12 kOe at 2 K [Fig. 4(c)]. Moreover, this metastable state can be completely reversed to the initial state after thermal treatment at 150 K or by illumination with a green 532 nm light, as a result, and it can work in repetitive cycles of 840 nm irradiation at 5 K and heating up to 150 K or 840 nm–532 nm excitation cycles at 5 K [Figs. 4(d) and 4(e)]. Finally, both critical temperatures and magnetic coercive field values for I magnet and II photomagnet in the metastable state are close to each other, implicating that the electron-transfer-coupled spin transition strongly depends on local geometry of Co(II) centers, but the strength of magnetic coupling is predefined by the structural parameters of the network, especially, the number of cyanido-bridges.

Since the discovery of [Co^II₃(pym)₄[W^V(CN)₈]₂ ⋅ 6H₂O coordination polymer revealing a photoinduced long-range magnetic ordering with a T_C of 40 K and H_C of 12 kOe at 2 K,^68–70 it seemed that it was impossible to design a material with better properties, but in science there are no impossible things, they are only unexplored. In 2012, novel 3D [{Co^II(4-Mepy)(pym)}₂{Co^II(H₂O)₂}{W^V(CN)₈}₂] ⋅ 4H₂O assembly (4-Mepy = 4-methylpyridine) has been described.⁷¹ The crystal structure of this compound consists of octacyanidotungstate(V) anions bridged to four [Co1/Co2(4-Mepy)(pym)]²⁺ and one [Co2(H₂O)₂]²⁺ complex cations. The four equatorial positions of Co1 (or Co2) are occupied by cyanide nitrogen atoms, while the apical positions are surrounded by two nitrogen atoms of 4-methylpyridine and bridging pyrimidine. The Co3 (or Co4) centers are coordinated by two cyanide nitrogen atoms, two pyrimidine nitrogen atoms, and two water oxygen atoms. It is noteworthy to mention that [{Co^II(4-Mepy)(pym)}₂{Co^II(H₂O)₂}{W^V(CN)₈}₂] ⋅ 4H₂O forms the cyanido-bridged folded square grids with four-metallic units along the ab-plane which are stabilized by Co1-pym-Co4, Co2-pym-Co3, CN-Co3-NC, and CN-Co4-NC bridges, and crystallization of water molecules filling the empty spaces in channels [Fig. 3(c)]. The closest interlayers distance adopts a value of 10.13 Å, and this is the shortest distance among known L–Co^II–[W^V(CN)₈] materials.

Magnetic studies supported by spectroscopic studies (IR and UV-vis spectroscopies) provided evidence that this system exhibits the electron-transfer-coupled spin transition effect from the HT phase Co^II–HS(S = ³/₂)—N=C—W^V(S = ¹/₂) to LT phase Co^III–LS(S = 0)—N=C—W^IV(S = 0) in the 172 K (T_1/2^{HT to LT})–241 K (T_1/2^{LT to HT}) range with a 69 K wide thermal hysteresis loop [Fig. 5(a)]. The width of this hysteresis is an intermediate value between 49 and 90 K hysteresis for Cs^I{[Co^II(3-CNpy)₂][W^V(CN)₈]} ⋅ H₂O and [Co^II₃(pym)₄[W^V(CN)₈]₂ ⋅ 6H₂O, respectively. More important aspect of this work is the observation of 785 nm diode-laser light irradiation, wavelength corresponding to a metal-to-metal charge transfer (MMCT) band from W^IV to Co^III, generated photoinduced spontaneous magnetization with a T_C of 48 K [Fig. 5(b)], and uniquely large magnetic hysteresis loop with H_C of 27 kOe at 2 K [Fig. 5(c)]. Additional feature of this material is thermal reversibility toward the initial LT state after heating up to 170 K. This implicates that [{Co^II(4-Mepy)(pym)}₂{Co^II(H₂O)₂}{W^V(CN)₈}₂] ⋅ 4H₂O can operate in repetitive cycles of 785 nm irradiation at 5 K and heating up to 170 K cycles [Fig. 5(d)]. Significant improvement in T_C and H_C in respect to the previous best systems [Co^II₃(pym)₄[W^V(CN)₈]₂ ⋅ 6H₂O (T_C = 40 K and H_C = 12 kOe at 2 K) has been addressed to the decrease in Co—N=C—W distances in the ab-plane, improving superexchange magnetic coupling, and to the increase in magnetocrystalline anisotropy. These improvements of magnetic characteristics are caused by the steric effect of 4-methylpyridine on the structure of channels.

Alternative route to control the temperatures of electron-transfer-coupled spin transition by humidity has been discovered for previously reported [{Co^II(4-Mepy)(pym)}₂{Co^II(H₂O)₂}{W^V(CN)₈}₂] ⋅ nH₂O compound.⁷² This 3D network system consists of the cyanido-bridged parallelogram-brick wall layers with 12-metallic units stabilized by pyrimidine and cyanide bridges [Fig. 3(c)], and around six crystallization and two coordination water molecules per structural unit at 60% relative humidity (RH). Humidity-dependent powder x-ray diffraction studies, thermogravimetric measurements, and IR spectroscopy performed in the 80%–5% RH range confirmed that as the humidity decreases, the amount of crystallization water molecules systematically diminishes from six molecules in the 80% RH down to around three and half molecules in the 5% RH. Additionally, desorption of the water reduces the hydrogen bonds network.

This has a direct impact on the results of magnetic measurements. Along with a reduction in relative humidity is growing strongly transition temperature from high temperature phase Co^II–HS(S = ³/₂)—N=C—W^V(S = ¹/₂) to low temperature one Co^III–LS(S = 0)—N=C—W^IV(S = 0), reaching extreme values between 147 K (100% RH) and 191 K (5% RH). Meanwhile, values of the reverse process during heating remain almost unchanged with the average temperature of 240 K. Consequently, the values of thermal hysteresis [ΔT ≡ (T_1/2^{LT to HT}–T_1/2^{HT to LT})/2] at 100%, 80%, 60%, 40%, 20%, and 5% RH are equal to 95, 87, 79, 68, 65, and 54 K in sequence (Fig. 6). Another important aspect considered in this work was the determination of thermodynamic parameters of transition enthalpy (ΔH) and transition entropy (ΔS) using differential scanning calorimetry (DSC) measurements and the mean-field model of phase transition [Slichter–Drickamer’s model: G = αΔH + γ(1 – α)α+T{R[α ln α + (1 – α)ln(1 – α)]–αΔS}, where α is the fraction of the HT phase, γ is the interaction parameter, and R is the gas constant] for sample at 0% RH, giving ΔH of 25.6 kJ mol⁻¹, ΔS of 116 J K⁻¹ mol⁻¹, and γ = 9.0 kJ mol⁻¹. Additionally, the spin entropy change ΔS_spin of 46 J K⁻¹ mol⁻¹ (ΔS_spin = RlnW, W is degeneracy and it equals to ²⁰⁴⁸/₈) has been calculated and compared with the experimental ΔS, which leads to conclusion that only 40% of total ΔS originates from spin state switching, while the remaining 60% was attributed to the phonon mode.

Presently, a rapid advancement of theoretical calculation methods for a variety of molecular materials exhibiting nontrivial magnetic properties is observed. Following this trend, a new 2-D (H₅O₂){[Co^III–LS(4-Brpy)₂][W^IV(CN)₈]} coordination polymer,⁷³ where 4-Brpy is 4-bromopyridine, has been investigated. This compound contains cyanido-bridged folded square grids with four-metallic units separated by charge-compensating Zündel cations (H₅O₂⁺) and neutral water molecules [Fig. 1(c)]. Each negatively charged monolayer consists of octacyanidotungstate(IV) anions linked by four cyanide to four cobalt(III) centers. Meanwhile, each Co(III) center is equatorial cyanido-bridged to four anions and axial coordinated by two nitrogen atoms of 4-bromopyridine molecules. The average interlayers distance adopts the value of 13.19 Å that is larger than the value of 12.42 Å for the analogous 2D Cs^I{[Co^II(3-CNpy)₂][W^V(CN)₈]} ⋅ H₂O system with smaller interlayer cation.

Moreover, the studied material is diamagnetic in the 2–390 K temperature range, indicating that the stable electronic state is Co^III–LS (S = 0)—W^IV(S = 0). This wide temperature range of the Co^III–LS–W^IV phase has not been increased so far. First-principles calculations showed that the bandgap consists of a W^IV valence band and a Co^III conduction band which explains stability of this state and indicates high possibility of occurrence of the photomagnetic effect. Consequently, this idea has been confirmed, a 785 nm CW laser light irradiation induced the electronic state transformation from Co^II–HS(S = ³/₂)—N=C—W^V(S = ¹/₂) stable state to Co^III–LS(S = 0)—N=C—W^IV(S = 0) metastable state and it leads to a spontaneous magnetization with a Curie temperature of 27 K and a coercive field of 2 kOe [Figs. 7(a) and 7(b)]. The photoinduced metastable state is annealed after heating up to 80 K. Finally, this system can work in 785 nm irradiation and heating cycles [Fig. 7(c)].

Sub-THz light resonating materials can find application as optical devices and absorbers for next-generation wireless communications, e.g., millimeter wave communication or beyond 5G communication, they will take an essential role in the era of big data and the Internet of Things (IoT). In response to this demand, Rb^I{[Co^II(3-CNpy)₂][W^V(CN)₈]} has been prepared.⁷⁴ This material is rubidium analog of the first photomagnetic Cs^I{[Co^II(3-CNpy)₂][W^V(CN)₈]} ⋅ H₂O compound. Their structure is built of cyanido-bridged folded square grids with four-metallic units separated by charge compensating Rb⁺ ions [Fig. 1(b)]. The interlayers’ distance is 12.14 Å, which is slightly smaller than the value of 12.42 Å determined for cesium analog.

From the point of view of magnetic properties, the Rb^I{[Co^II(3-CNpy)₂][W^V(CN)₈]} exhibits electron-transfer-coupled spin transition with slightly narrowed thermal hysteresis loop of 40 K and with lower transition temperatures of T_1/2^{HT to LT} = 150 K and T_1/2^{LT to HT} = 190 K relative to the cesium assembly [Fig. 8(a)]. Besides, photomagnetic studies with 785 nm irradiation revealed a spontaneous magnetization with a T_C of 20 K [Fig. 8(b)] and magnetic hysteresis loop with H_C of 4.2 kOe at 2 K. The photoinduced metastable state can be completely reversed to the initial state after thermal treatment at 120 K.

The most important part of this work was the experimental detection (using terahertz time domain spectroscopy—THz-TDS) and theoretical calculation (by means the first-principle phonon mode calculations) of optical phonon mode spectra for LT, HT, and photoexcited metastable phases. The THz absorption spectrum of the HT phase has eight absorption peaks: 0.58, 0.78, 1.06, 1.22, 1.30, 1.41, 1.51, and 1.89 THz, and the LT phase has eight absorption peaks: 0.69, 0.87, 0.91, 1.14, 1.31, 1.42, 1.58, and 1.75 THz [Fig. 9(a)]. The first four peaks are assigned to the phonon modes of Rb⁺ vibrations in the ab-plane and along the crystallographic c axis. Meanwhile, other peaks correspond to a combination of transverse translational and librational modes of W–CN–Co and 3-cyanopyridine ligand rotation. Temperature-dependent THz-TDS supported by single crystal x-ray diffraction measurements confirmed a temperature-induced switching of low-frequency phonon modes of Rb^I{[Co^II(3-CNpy)₂][W^V(CN)₈]} with thermal hysteresis [Fig. 9(b)], identical to observed one in the magnetic measurement, and the frequency shifts of the optical phonon modes are triggered by a stronger Rb⁺ trapping force because of the contraction of the Rb-N lengths. The last part of this work was the discovery of photo-induced switching of the low-frequency phonon mode in the THz-TDS observed as 785 nm light induced conversion of the LT phase spectrum to the photoinduced metastable state with spectrum similar to the HT phase. This photo-induced switching effect is repetitively noted by light irradiation and thermal annealing at 120 K. Concluding, this is the first work investigating external-stimulation-controllable sub-terahertz (sub-THz) phonon crystals.

The first challenge to be faced is the development of new high-dimensional cyanido-bridged cobalt-octacyanidotungstate networks showing the electron-transfer-coupled spin transition phenomenon using diverse synthetic approaches. However, before setting new directions in the design of this class of compounds, it is necessary to summarize recent knowledge. So far all electron-transfer-coupled spin transition systems based on 2D cyanido-bridged folded square grids [M^I{[Co^II(3-CNpy)₂][W^V(CN)₈]} ⋅ xH₂O, M^I= Rb and Cs, and (H₅O₂){[Co^III–LS(4-Brpy)₂][W^IV(CN)₈]}]^67,73,74 and 3D networks with 12-metallic channels [[Co^II₃(pym)₄[W^V(CN)₈]₂ ⋅ nH₂O and [{Co^II(4-Mepy)(pym)}₂{Co^II(H₂O)₂}{W^V(CN)₈}₂] ⋅ nH₂O].^68–72 Besides, we can list out other 2D and 3D Co–W systems showing only long-range magnetic ordering and broad hysteresis loops at 2 K. Among them, two chiral magnets, based on enantiopure (R)-1-(4-pyridyl)ethanol) [R-1-(4pyEtOH)], can be recognized: 3D cyanido-bridged network with 12-metallic channels [Co^II(H₂O)]₂{Co^III[μ-R-1-(4pyEtOH)]₂}[W^IV(CN)₈][W^V(CN)₈] ⋅ 5H₂O¹¹⁹ and 3D cyanido-bridged network with square-shaped columns linked by pyrimidine and CN⁻ linkers [Co^II(H₂O)₂]{Co^II(pym)[μ-R-1-(4pyEtOH)]}₂[W^V(CN)₈]₂ ⋅ 7.5H₂O.¹²⁰ Moreover, both systems characterize very similar magnetic behavior with a long-range ferromagnetic ordering with Curie temperature (T_C) of 11 and 19 K, respectively, and the coercive fields (H_C) at 2 K of 1.5 and 3.6 kOe, respectively. The decrease in T_C and H_C values between them originates from the smaller number of cyanide bridges involved in the long-range ferromagnetic coupling through —Co^II–HS(S = ³/₂)—N=C—W^V(S = ¹/₂)—C=N—Co^II–HS(S = ³/₂)— connectors and lower dimensionality of magnetic interaction networks. Another example is 3D hybrid inorganic organic magnet {[Co^II(H₂O)₂]₂[Co^II₂(H₂O)₂(μ-C₂O₄)][W^V(CN)₈]₂ ⋅ 6H₂O}_n, built of 2D cyanido-bridged layers pillared by oxalate-bridges, exhibiting T_C of 26 K and H_C of 3.0 kOe.¹²¹ The last two systems are 3D cyanido-bridged honeycomb network with 12-metallic units Co^II₃[W^V(CN)₈]₂(pu)₂ ⋅ nH₂O (pu = purine)¹²² and 3D cyanido-bridged network with 16-metallic units [Co^II(H₂O)_1.5(Cl)_0.5]₄[Co^II(H₂O)₂]₃[W^V(CN)₈]₄ ⋅ 17H₂O.¹²³ The first assembly shows humidity-induced switching between two distinct ferromagnetic phases with T_C of 49 and 29 K for low humidity (n = 4.3) and high humidity (n = 8.5) phases, respectively, and magnetic hysteresis loops with coercive fields of 6.2 and 0.6 kOe, respectively. The second product exhibits ferromagnetic ordering with T_C of 29 K and H_C of 5.5 kOe.

As it is shown in aforementioned works, nontrivial magnetic behaviors can be mainly achieved for coordination polymers with planar terminal ligands [i.e., 3-cyanopyridine, 4-bromopyridine, 4-metylpyridine, and 1-(4-pyridyl)ethanol]) or bridging aromatic ligands (i.e., pyrimidine and purine); thus, it is obvious that analogous ligands with diverse functional group should be promising. Special consideration should be paid to organic ligands bearing intrinsic functionalities such as chirality (e.g., containing asymmetric carbon atoms), conductivity (e.g., aromatic amino acids forming zwitterions), or luminescence (e.g., aromatic carboxylic acids). Another important aspect worth considering is the replacement of cesium and rubidium salts in the syntheses by potassium, sodium, or lithium cations to enhance conducting properties or generate new kinds of responses in THz absorption spectroscopy. Last but not least, the partial substitution of cobalt(II) with another 3d block metals [e.g., Mn(II), Fe(II), or Ni(II)] should be also attractive considering that this modification has been successfully applied for trimetallic {Co^II_9-xM^II_x[W^V(CN)₈]₆(MeOH)₂₄}⋅MeOH (M^II = Mn, Fe, and Ni) clusters. In the case of Co^II–Fe^II–W^V systems, two cooperating electron-transfer channels, Co^II–HS(S = ³/₂)—N=C—W^V(S = ¹/₂) (HT) $→←$ Co^III–LS(S = 0)—N=C—W^IV(S = 0) (LT) and Fe^II–HS(S = 2)—N=C—W^V(S = ¹/₂) (HT) $→←$ Fe^III–HS(S = ⁵/₂)—N=C—W^IV(S = 0) (LT) have been detected.^124,125 Meanwhile, the substitution with manganese and nickel ions amplified anisotropy of cobalt centers resulting in slow magnetic relaxation at low temperature phases characteristic for single molecule magnets.¹²⁶ 

Advances in research sometimes require us to return to the beginning and look at known materials from a different point of view. With this idea in mind, it is worth looking at potential functionalities that were not originally explored and which may accelerate the development of other fields of science and set a new path toward marvellous discoveries. A perfect confirmation of this concept is the observation of temperature- and photo-induced switching of low-frequency phonon modes of Rb^I{[Co^II(3-CNpy)₂][W^V(CN)₈]} crystallites. With the development of THz-TDS technology, the established material finds novel applications that may be more significant for the development of high-speed wireless communication than the originally considered molecular magnetism. At this point, it is worth highlighting that caesium(I) analog should be even more promising due to the fact that heavier alkali cation should shift phonon modes to lower energy of the sub-terahertz region, which is now the subject of further research. Another recently investigated aspect, related to the research on layered systems separated by small ions, is their potential to exhibit superionic conductivity, which can be used in the construction of batteries and supercapacitors. This kind of property is especially expected for (H₅O₂){[Co^III–LS(4-Brpy)₂][W^IV(CN)₈]} which comprise Zündel ions well known in molecular materials to exhibit a high proton conductivity via the hydrogen-hopping mechanism.^127–131 Furthermore, this assembly is promising candidate for electrochemical-induced spin transition systems whose concept is based on previous electrochemical studies for Co^III/II–W^IV/V couples.¹³² Finally, the influence of humidity, desolvation, and the presence of other solvents on 3D networks with 12-metallic channels filled with water [{Co^II(4-Mepy)(pym)}₂{Co^II(H₂O)₂}{W^V(CN)₈}₂] ⋅ 4H₂O and [Co^II₃(pym)₄[W^V(CN)₈]₂ ⋅ 6H₂O is also strongly considered. Already presented humidity-controlled thermal phase transition in the first compound and several other porous molecular magnets, molecular magnetic sponges, and solvatomagnets confirm the validity of this study and the demand for their further development. Therefore, we can expect a strong increase in the number of articles related to this research in the near future.

In addition to the methods of obtaining new functionalities from known materials, it is also worth mentioning an equally important aspect, which is broadening the knowledge about the spin transition mechanism in L–Co^II–[W^V(CN)₈] systems. Originally, this mechanism was intuitively attributed to a charge-transfer-induced spin transition (CTIST) from high temperature phase Co^II–HS(S = ³/₂)—N=C—W^V(S = ¹/₂) to low temperature one Co^III–LS(S = 0)—N=C—W^IV(S = 0).⁶⁷ However, in light of the current results obtained using femtosecond x-ray and optical absorption spectroscopies for cobalt(II)-hexacyanidoferrate(III) Prussian blue analogs,⁶¹ the second mechanism, based on successive the spin transition on Co center and the associated structural reorganizations drive the charge transfer, can be postulated and requires confirmation. This issue will be another challenge for the coming years and, at the same time, a milestone in the development of new materials.

Summing up this Perspective, it should be noted that it was possible to obtain many unique physicochemical effects (e.g., non-trivial magnetic properties—spin transition phenomenon and photomagnetism, millimeter wave absorption) for few high-dimensional Co–[W(CN)₈] systems. They left a permanent mark on science, opening us the way to new discoveries. We also believed that this Perspective will guide and motivate researchers to face up to the challenges of further development of this research field.

This Perspective was supported by the Japan Society for the Promotion of Science within the Grant-in-Aid for Specially Promoted Research (Grant No. 15H05697), Grants-in-Aid for Scientific Research (A) (Grant No. 20H00369), and the Grants-in-Aid for Scientific Research on Innovative Areas Soft Crystals (Area No. 2903, Grant No. 17H06367). This Perspective was partially supported by MEXT Quantum Leap Flagship Program (MEXT Q-LEAP) under Grant No. JPMXS0118068681. We also acknowledge The University of Tokyo Cryogenic Research Center and Center for Nanolithography and Analysis, which are supported by MEXT, and IM-LED LIA (CNRS).

The data that support the findings of this study are available from the corresponding author upon reasonable request.