Dynamic functions of bis- and tris(saloph) cobalt(III) structures based on axial coordination

Abstract

This review article focuses on the functionalization and dynamic functional switching of low-spin d⁶ cobalt(III) complexes derived from various oligo(saloph) structures, such as bis(saloph) macrocycles and tris(saloph) cage complexes (H₂saloph = N,N′-disalicylidene-o-phenylenediamine). The bis(saloph) dicobalt(III) complexes with methylene- or phenylene-bridged ligands exhibit reversible redox-driven structural switching, in which the axial functional ligands dissociate and reassociate in response to the Co^III/Co^II interconversion. The ether-bridged macrocyclic bis(saloph) cobalt(III) complexes show excellent cation binding affinity at the central O₆ binding site, which is significantly influenced by the nature of the axial ligands at the cobalt centers. In particular, an anion-capped structure leads to the formation of a unique metastable host–guest complex, enabling stimuli-responsive behavior upon external triggering. Post-metalation ligand exchange with anionic ligands and bridging diamine ligands provides a versatile strategy for structural and functional tuning of these macrocyclic hosts. In some complexes, the ligand exchange reactivity and the guest binding affinity enhance each other. A helical tris(saloph) cobalt(III) cryptand exhibits dynamic P/M chirality interconversion via the axial ligand exchange involving achiral or chiral amines, allowing precise control over the chirality inversion rates and enabling a unique transient chirality inversion during racemization. Furthermore, closed-cage metallocryptands bearing bridging diamine ligands effectively suppress the guest uptake/release kinetics. Thus, the introduction, removal, and exchange of axial ligands (X) in the [Co(saloph)X₂]⁺-type units have been successfully employed for the functionalization and dynamic switching of metallohosts and metallo-supramolecular structures.

---

Shigehisa Akine

Shigehisa Akine received his Ph.D. in 2000 from the University of Tokyo under the supervision of Prof. Takayuki Kawashima, then worked as a research associate with Prof. Tatsuya Nabeshima at the University of Tsukuba. He became assistant professor in 2004 and associate professor in 2008. In 2013, he joined Kanazawa University as a full professor and in 2017 became a principal investigator at the Nano Life Science Institute (WPI-NanoLSI). His research focuses on kinetic control of structural conversions in responsive molecules, especially in host–guest, supramolecular, coordination, and organic chemistry.

---

1. Introduction

The salen ligand†¹ and its phenylene analog, the saloph ligand,‡¹ are versatile tetradentate chelate ligands that provide a planar N₂O₂ coordination environment (Fig. 1a). Some of the resulting transition metal complexes have been used as catalysts,² liquid crystals,³ and sensors,⁴ and are known to exhibit unique magnetic⁵ and optical properties,⁶ as well as biological activities.⁷ In addition to simple monomeric structures, many oligomeric compounds containing multiple salen units have been developed,^8,9 and their material applications¹⁰ and catalytic functions¹¹ have been studied. These include macrocyclic oligomers,^8,9 as well as acyclic oligomers¹² and molecular cages.¹³ In particular, cyclic oligomers and cage structures are advantageous because they can be obtained in relatively high yields, thanks to the dynamic nature of the imine C=N double bonds.¹⁴ Some of these compounds exhibit unique host–guest binding affinities toward molecular and ionic guest species,⁹ taking advantage of the rigid and shape-persistent nature of the salen/saloph complex structures, which arises from their well-defined chelate coordination motifs.

Fig. 1 (a) Chemical structures of H₂salen, H₂saloph, and [Co(saloph)X₂]⁺. Functional groups can be introduced as axial ligands X. (b) X/Y axial ligand exchange in [Co(saloph)X₂]⁺ structures, allowing for tunable properties at the axial positions.

Among the metals that can be incorporated into these oligo(salen)/-(saloph) structures for metallohosts and metallosupramolecular systems, those with a d⁸ electron configuration, such as nickel(II)¹⁵ and palladium(II),¹⁶ are particularly advantageous. Their square-planar geometry aligns well with the planar N₂O₂ coordination environment provided by the salen/saloph ligands. Metalation of these ligands yields rigid square planar complexes through the simultaneous formation of four coordination bonds, which facilitates the predictable and selective formation of shape-persistent structures. In addition, the resulting d⁸ square-planar complexes are diamagnetic. The use of diamagnetic metals is, in fact, essential for the investigation of metal-containing supramolecular and host–guest structures, as it allows for NMR measurements for structural elucidation and detailed analysis of host–guest binding.

In contrast to the square-planar metal ions, pentacoordinate metal centers situated in the salen/saloph coordination pocket can accommodate an additional monodentate ligand. For example, parent salen/saloph ligands are known to form mononuclear pentacoordinate complexes formulated as [Zn(salen)X]/[Zn(saloph)X] (X = H₂O, py, etc.)¹⁷ or dimeric complexes [Zn₂(salen)₂]/[Zn₂(saloph)₂] in which one of the phenoxo oxygen atoms occupies the apical position of the counterpart Zn²⁺.¹⁸ This fifth coordination has been utilized for the construction, structural conversion, and functionalization of various types of multi-metal self-assembled structures.^19–21

Analogously, hexacoordinate complexes, [M(salen)X₂]/[M(saloph)X₂], can be obtained by introducing two additional monodentate ligands (X) to the metal centers in the salen/saloph coordination site. In fact, the salen/saloph ligands can accommodate various metal ions with an octahedral geometry, which is the most common and ubiquitous coordination structure adopted by a wide range of transition metal ions. In most cases, the N₂O₂ donor set of the salen/saloph ligands occupies four equatorial positions around the octahedral metal ion. Accordingly, the two X ligands in the [M(salen)X₂]/[M(saloph)X₂] complexes occupy trans positions to each other, located at the axial positions relative to the MN₂O₂ plane in the [M(salen)]/[M(saloph)] structures (Fig. 1a).²²

In particular, among various salen/saloph complexes containing an octahedral metal ion, the cobalt(III) complexes, [Co(salen)X₂]⁺/[Co(saloph)X₂]⁺,^23,24 offer significant advantages. Owing to the large ligand field splitting originating from the low-spin d⁶ electron configuration, diamagnetic complexes are usually obtained exclusively and predictably, which facilitates investigation based on NMR spectroscopy. In addition, these low-spin cobalt(III) complexes are generally inert and stable, which allows various types of site-selective functionalizations at the axial positions without loss of the central cobalt(III) ion from the salen/saloph structures. Ligand exchange occurs slowly and only at the two axial X positions in the [Co(salen)X₂]⁺/[Co(saloph)X₂]⁺ structures, on a timescale of minutes to hours,²⁵ owing to their inert nature (Fig. 1b). This reactivity is useful for the slow, time-dependent control of functions in the multi-metal structures.²⁶ Although kinetic inertness often hampers the integration of dynamic functions into metal complexes, [Co(salen)]/[Co(saloph)] structures offer a distinct advantage: their axial positions remain sufficiently reactive while the equatorial CoN₂O₂ core maintains structural integrity. This spatially controlled reactivity enables the rational design of dynamic, switchable systems based on inert cobalt(III) centers.

Cobalt(III) complexes are particularly suited for functionally relevant structural transformations via selective axial ligand exchange, even though other transition metal complexes such as Pd, Pt, and Fe also exhibit kinetic inertness. Furthermore, the redox activity of cobalt(III), especially the Co^III/Co^II couple, offers additional opportunities for external control and switching of functions,²⁷ which is less accessible in Pd, Pt, Zn, or Fe systems. Indeed, a wide range of cobalt(III)-based redox-responsive functional molecules and supramolecular architectures have been developed, demonstrating the broad versatility and growing importance of switchable metal-containing structures.²⁸

This review article focuses on the functionalization and dynamic switching of the low-spin d⁶ cobalt(III) complexes derived from various kinds of oligo(saloph) structures, such as bis(saloph) ligands (H₄L¹, H₄L², H₄L³, H₄L⁴) and the tris(saloph) cage (H₆L⁵) (Fig. 2). In particular, introduction, removal, and exchange of the axial ligands X in the [Co(salen)X₂]⁺/[Co(saloph)X₂]⁺ structures²⁶ have been efficiently employed for functionalization and the dynamic functional switching of metallohosts and metallo-supramolecular structures incorporating cobalt(III) centers.

Fig. 2 Chemical structures of bis(saloph) compounds, H₄L¹, H₄L², H₄L³, H₄L⁴, and tris(saloph) cage H₆L⁵.

2. Functions of bis(saloph) cobalt(III) complexes based on axial coordination

2.1. Methylene-bridged macrocyclic bis(saloph) dicobalt(III) complexes for redox-driven structural switching

A bis(saloph) macrocyclic ligand, H₄L¹ (Fig. 2),²⁹ in which two H₂saloph motifs are connected by two methylene linkers to form a macrocyclic framework, was synthesized. A series of doubly bridged dicobalt(III) complexes, L¹Co₂(DAn)₂, was prepared by reacting H₄L¹ with cobalt(II) acetate in the presence of various diamine ligands (DAn = DA1–DA6) under aerobic conditions (Fig. 3a).³⁰ These complexes were characterized by spectroscopic methods and X-ray crystallography, revealing dinuclear structures featuring two diamagnetic cobalt(III) ions bridged by diamine ligands, such as alkanediamines DA1–DA4 with varying methylene chain lengths (Fig. 3a). The use of oligo(ether) diamines, DA5 and DA6, introduced potential cation binding sites into the dicobalt(III) macrocycle, analogous to crown ethers.

Fig. 3 Redox switching of dinuclear complexes L¹Co₂(DAn)₂. (a) The oxidized state with octahedral cobalt(III) ions. (b) The reduced state with tetracoordinate cobalt(II) ions.

While the cobalt(III) ions in the L¹Co₂(DAn)₂ complex prefer a hexacoordinate octahedral geometry (Fig. 3a), reduction to cobalt(II) leads to a less coordinated geometry (Fig. 3b). This Co^III/Co^II redox transformation can be exploited to switch between the bound and unbound states of the axial ligands at the cobalt centers. Electrochemical measurements exhibited redox waves with large peak separations, suggesting that the redox processes are accompanied by significant changes in coordination geometry.³⁰ Specifically, the Co^III → Co^II reduction occurs in the hexacoordinate state with bridging diamine ligands, whereas the Co^II → Co^III oxidation occurs in a tetracoordinate state lacking the diamine ligands (Fig. 3b). The reversible structural transformation between these two states was further confirmed by electrolytic absorption spectroscopy as well as mass spectrometry.

Thus, the doubly bridged structure can be constructed and destructed by association/dissociation of the axially coordinating diamine ligands DAn in response to redox changes. This would significantly change the host–guest binding behavior in the polyether-based cavities of L¹Co₂(DA5)₂ and L¹Co₂(DA6)₂. However, these dinuclear cobalt(III) complexes did not exhibit any binding affinity for alkali metal ions such as Li⁺, Na⁺, or K⁺.²⁹ This may be attributed to the positive charge of the dicationic L¹Co₂ core, which experiences strong electrostatic repulsion with cationic guests.

2.2. Phenylene-bridged bis(saloph) dicobalt(III) complexes for redox-driven structural switching

The acyclic bis(saloph) ligand, H₄L² (Fig. 2), in which two H₂saloph units share a single phenylenediamine subunit, is known to form a series of dinuclear complexes where the two metal centers are electronically coupled.^20,24,31 This H₄L² ligand was used to prepare the dinuclear cobalt(III) complex L²Co₂(B1)₄, which features four crown ether subunits at the axial positions of the cobalt(III) centers (Fig. 4).³² The complex was synthesized by the reaction of H₄L² with cobalt(II) acetate in the presence of 4-aminomethylbenzo-15-crown-5 (B1) under aerobic conditions.

Fig. 4 Redox-driven structural conversion of crown-ether-functionalized complex L²Co₂(B1)₄. (a) Oxidized state and (b) reduced state. Only the oxidized form L²Co₂(B1)₄ can bind a K⁺ ion in a sandwich fashion, enabling redox-switchable guest binding.

Since each face of the L²Co₂(B1)₄ complex functions as a bis(15-crown-5) host capable of binding a cationic guest in a sandwich fashion, the complex was expected to bind two K⁺ ions on both faces to form a 1 : 2 (host/guest) complex. However, spectroscopic measurements revealed that this L²Co₂(B1)₄ complex exhibits K⁺ binding with 1 : 1 stoichiometry, as confirmed by Job plot analysis and mass spectrometry. This binding behavior can be rationalized by the molecular deformation caused by the first guest binding in a [(15-crown-5)₂K]⁺ sandwich fashion, which may increase the distance between the two crown ether moieties on the opposite face, thereby suppressing the second K⁺ binding. This L²Co₂(B1)₄ complex also undergoes a reversible Co^III/Co^II redox interconversion, accompanied by the association/dissociation of the axially coordinating crown ether subunits B1 (Fig. 4).³²

The Co^III/Co^II redox interconversion of the same dinuclear L²Co₂ motif was also exploited for the reversible association/dissociation of dendrimer subunits G3. The dinuclear complex L²Co₂(G3)₄, which contains four dendrimer subunits G3 (Fig. 5),³³ was expected to undergo Co^III/Co^II redox switching accompanied by the reversible binding of the dendrimer subunits G3. However, under electrochemical conditions, L²Co₂(G3)₄ did not exhibit efficient redox switching, likely due to the steric bulkiness of the dendrimer subunits G3. Instead, photo-driven redox reactions proved effective: upon photoirradiation (λ ≥ 420 nm) in a degassed DMF solution containing triethanolamine as a sacrificial electron donor, the complex was reduced to a tetracoordinate dicobalt(II) species. This reduced form can be re-oxidized by air exposure to regenerate the initial hexacoordinate dicobalt(III) complex. Thus, despite the steric hindrance from the bulky dendrimer subunits, reversible association/dissociation of the dendrimer subunits in L²Co₂(G3)₄ was successfully achieved via photo-driven redox switching.

Fig. 5 Chemical structure of the dendrimer-functionalized dinuclear complex L²Co₂(G3)₄.

2.3. Ether-bridged bis(saloph) dicobalt(III) macrocycles for controlled guest binding

An ether-bridged bis(saloph) macrocycle, H₄L³ (Fig. 2), which is an oxygen analogue of the methylene-bridged bis(saloph) macrocycle H₄L¹, was synthesized.³⁴ Owing to the ether linkages, the metalated form, L³M₂, features an 18-crown-6-like central binding cavity surrounded by six oxygen donor atoms, which exhibit excellent binding affinity for cationic guest species. For example, the nickel(II) metallohost L³Ni₂ (Fig. 6a) strongly binds to Na⁺ to form a 1 : 1 host–guest complex and interacts with larger alkali metal ions (K⁺, Rb⁺, Cs⁺) to afford unique stacked structures.^34–38 This higher binding affinity arises from the combination of negatively polarized phenoxo oxygen atoms and the well pre-organized structure of the metallohosts. This section focuses on the guest binding behavior of the hexacoordinate cobalt(III) analogues L³Co₂A₄ (Fig. 6b), which have four primary amine ligands A at the axial positions of the cobalt(III) centers.

Fig. 6 Chemical structures of (a) L³Ni₂ and (b) L³Co₂A₄.

Since the dinuclear cobalt(III) metallohosts L³Co₂A₄ are dicationic, they were expected to show a poor cation binding affinity due to electrostatic repulsion, especially in comparison to the non-charged nickel(II) analogue L³Ni₂. Contrary to expectations, however, the cobalt(III) metallohost L³Co₂(MeNH₂)₄, bearing four methylamine ligands, showed excellent binding affinity toward various cationic guests, such as a monovalent cation, Na⁺ (K_a = 8.5 × 10⁶ M⁻¹), and even a multivalent cation, La³⁺ (K_a = 2.4 × 10⁶ M⁻¹).³⁹ In the crystal structure of the Na⁺ inclusion complex, L³Co₂(MeNH₂)₄Na, the Na⁺ ion is located precisely at the center of the O₆ binding site (Fig. 7a). The triflate counter anions not only directly coordinate to the Na⁺ ion but also form hydrogen bonds with the methylamine NH₂ groups on both faces of the macrocycle, resulting in a unique anion-capped structure (Fig. 7b).³⁹ These noncovalent interactions appear to contribute not only to the strong cation binding, but also to blocking guest entry/exit, acting like a cap or lid on a container.

Fig. 7 (a) X-ray crystal structure of L³Co₂(MeNH₂)₄Na. The two TfO⁻ counter anions are also shown. (b) Schematic drawing of the anion-capped structure of the host–guest complex, L³Co₂(MeNH₂)₄Na. (c) Plots of mole fractions of the La³⁺ complex versus time after the addition of K⁺ in CD₃OD: (i) in the absence of AcO⁻; (ii) in the presence of AcO⁻; (iii) guest exchange was initiated when AcO⁻ was added as a trigger after 120 h. (d) Conversion of the metastable state [K⁺ complex + unbound La³⁺] into the thermodynamically most stable state [La³⁺ complex + unbound K⁺] triggered by AcO⁻. Adapted in part with permission from ref. 39.

In fact, the guest uptake/release rates of the metallohost L³Co₂(MeNH₂)₄ were found to be slow on the ¹H NMR time scale for cationic guests such as Na⁺, K⁺, and Ca²⁺. Notably, the uptake of the La³⁺ ion was particularly slow (k_in ≈ 10⁻² M⁻¹ s⁻¹), requiring nearly 100 h to reach completion. This remarkably slow uptake was attributed to the anion-capped structure, as the uptake rate was found to depend significantly on the nature of counter anions.

This anion-capped structure also contributed to the deceleration of guest exchange in the metallohost L³Co₂(MeNH₂)₄. A compelling demonstration is the formation of a metastable host–guest complex, that is, a kinetically trapped state in which a weaker guest is preferentially bound within the host cavity, leaving a stronger guest unbound. From a thermodynamic viewpoint, the metallohost showed a lower affinity for K⁺ (K_a = 1.1 × 10⁶ M⁻¹) than for La³⁺ (K_a = 2.4 × 10⁶ M⁻¹). However, when K⁺ and La³⁺ were simultaneously added, K⁺ was selectively taken up for kinetic reasons. Guest exchange to the thermodynamically favored La³⁺ complex was not observed even after 2 weeks (Fig. 7c(i)), indicating that the guest exchange was almost completely kinetically suppressed. Typically, guest binding in simple crown ethers is fast enough that the thermodynamically most stable complex is always selectively formed. In this context, the cobalt(III) metallohost L³Co₂(MeNH₂)₄ gives the first metastable host–guest complex derived from a simple macrocyclic host (Fig. 7d).^39,40

Such a metastable state retains the potential to transition to the thermodynamically most stable structure at any time when triggered. In the aforementioned case, the mixture of the K⁺ complex with unbound La³⁺ remained in a kinetically trapped metastable state, where the conversion to the thermodynamically most stable state was suppressed, but this conversion was accelerated by the addition of acetate ion as a trigger (Fig. 7c(ii), (iii) and d). This represents a new type of on-demand, stimuli-responsive function that exploits the long-lived yet transformable nature of a metastable host–guest complex.

Thus, the guest uptake/release rates of the metallohost L³Co₂(MeNH₂)₄ were found to be significantly decelerated by its anion-capped structure. These rates were also expected to be influenced by the structure of the amine ligands A coordinating to the cobalt(III) ions. In order to clarify this effect, a series of dinuclear metallohosts L³Co₂A₄, each bearing four primary monoamine ligands (A = EtNH₂, PhCH₂NH₂, PhC₂H₄NH₂, PhC₃H₆NH₂) with or without a phenyl group remote from the O₆ binding cavity, were synthesized (Fig. 6b). All these metallohosts exhibited a consistent selectivity trend among alkali metal ions: Na⁺ > K⁺ > Rb⁺ > Cs⁺. This trend was primarily attributed to differences of up to ∼500 000-fold in the release rate constants k_out, following the order of Na⁺ < K⁺ < Rb⁺ < Cs⁺.⁴¹

The structural variation in the amine ligands A also affected the binding behavior in both thermodynamic and kinetic aspects (Table 1). For example, the binding constants of L³Co₂A₄ (A = EtNH₂, PhCH₂NH₂, PhC₂H₄NH₂, PhC₃H₆NH₂) with Na⁺ differed by up to 200-fold. The introduction of a phenyl group generally weakened the binding, with the benzylamine analogue, L³Co₂(PhCH₂NH₂)₄, showing the lowest affinity. A detailed analysis of the uptake/release rates, k_in and k_out, revealed that the phenyl-containing derivatives generally exhibited a faster release rate k_out and slower uptake rate k_in. Since the Na⁺-bound species of these phenyl-containing complexes, L³Co₂(PhC_nH_2nNH₂)₄Na (n = 1,2,3), have quite similar structures to each other, in which the Na⁺ guest is located at the center of the O₆ cavity, the differences in the binding constants K_a are mainly ascribed to variations in the uptake rates k_in. Crystallographic analysis of the guest-free metallohosts, L³Co₂(PhC_nH_2nNH₂)₄ (n = 1,2,3), revealed that some of the phenyl and methylene C–H groups interact with the oxygen atoms of the O₆ binding site via C–H⋯O interactions. This suggests that guest binding requires additional energy to overcome this extra stabilization (Fig. 8).⁴¹

Fig. 8 Guest binding equilibrium of L³Co₂(PhC_nH_2nNH₂)₄ (n = 1,2,3). The definitions of the uptake and release kinetic rate constants, k_in and k_out, for the host–guest binding equilibrium are shown. Extra stabilization via C–H⋯O interactions by the phenyl capping can explain the relatively weaker binding compared to L³Co₂(EtNH₂)₄.

Table 1 Binding constants and uptake/release rate constants for the host–guest complexes of L³Co₂A₄ with Na^+ a,b

	Ka (M^−1)	kin (M^−1s^−1)	kout (s^−1)
a In CD3OD.b Determined by ^1H NMR spectroscopy. Data taken from ref. 41.
A = EtNH2	9.6 × 10^6	1.1 × 10^6	0.12
A = PhCH2NH2	4.9 × 10^4	5.2 × 10^4	1.05
A = PhC2H4NH2	3.8 × 10^5	9.0 × 10^4	0.23
A = PhC3H6NH2	8.1 × 10^5	1.0 × 10^5	0.12

---

2.4. Ligand exchange of bis(saloph) cobalt(III) metallohosts for control of the guest binding affinity

Since cobalt(III) is inert due to its low-spin d⁶ electron configuration, dissociation of the cobalt(III) ion from the tetradentate saloph chelate ligand is essentially negligible. Nevertheless, the Co–N bonds to the monoamine ligands A are relatively dynamic, thus allowing the ligand exchange in a site-selective fashion (Fig. 1b). This reactivity is advantageous for functionalization after the L³Co₂ dinuclear structures are constructed, i.e., the post-metalation modification enables alteration or fine-tuning of molecular functions, such as the host–guest binding affinity, as seen in the post-synthetic modification of various molecular cages and MOFs.⁴² Indeed, the guest binding affinity of the dinuclear metallohosts L³Co₂A₄ was significantly influenced by the structure of the primary amine ligands A introduced at the cobalt(III) centers, as described in the previous section.

In this context, the synthesis of analogous L³Co₂A₄ complexes bearing different types of amine ligands, such as a secondary amine (pip = piperidine) and a tertiary amine (quin = quinuclidine), was attempted using the same protocol as that for L³Co₂(MeNH₂)₄.⁴³ Whereas the piperidine-coordinating complex, L³Co₂(pip)₄, was successfully obtained, the quinuclidine-coordinating complex, L³Co₂(quin)₄, was not obtained.

Regarding the ligand exchange reactivity, the methylamine ligands in L³Co₂(MeNH₂)₄ were not readily exchanged with other amines (pip, quin) (Fig. 9a). In contrast, the piperidine-coordinating complex, L³Co₂(pip)₄, was efficiently converted into the methylamine-coordinating complex, L³Co₂(MeNH₂)₄ (Fig. 9b), but not into the quinuclidine complex, L³Co₂(quin)₄ (Fig. 9c). Thus, the affinity order of the amines for the cobalt(III) centers in L³Co₂A₄ followed the trend of primary amine > secondary amine > tertiary amine, which can be attributed to the steric bulk around the nitrogen donor atom. Among the isolable L³Co₂A₄ complexes, the piperidine-coordinating complex, L³Co₂(pip)₄, was found to be the best starting complex for ligand exchange, as it exhibited the highest reactivity.⁴³

Fig. 9 Structural conversion of L³Co₂(pip)₄ and related complexes via ligand exchange. Path (h) represents the acceleration of ligand exchange with AcO⁻ in the presence of Na⁺.

It is noteworthy that the ligand exchange of L³Co₂(pip)₄ with pyridine selectively afforded a di-exchanged product, L³Co₂(pip)₂(py)₂, instead of the tetra-exchanged species, L³Co₂(py)₄ (Fig. 9d). X-ray crystallography and ¹H NMR spectroscopy revealed the anti-diagonal stereoconfiguration of this product, in which each cobalt center bears one pip and one py ligand occupying opposite positions on the two faces, e.g., pip above and py below on one cobalt center, and reverse on the other. This stereoselectivity can be attributed to the doubly curved structure of the L³Co₂ framework, in which the bulkier piperidine ligands preferentially occupy the two convex faces, while the pyridine ligands reside on the narrower concave faces. The same product, the anti-diagonal L³Co₂(pip)₂(py)₂, was also selectively formed via the reverse ligand exchange (py → pip) starting from L³Co₂(py)₄ (Fig. 9e).

A similar anti-diagonal di-exchanged complex, L³Co₂(pip)₂(OAc)₂, was obtained by the reaction of L³Co₂(pip)₄ with acetate ion (Fig. 9f). This complex was obtained as precipitates directly from the reaction mixture in 82% isolated yield.

These results demonstrate that a variety of post-metalation modifications of L³Co₂A₄ metallohosts can be achieved by exploiting the reactivity of the axial ligands coordinating to the cobalt(III) centers.⁴³ Such modulations provide a means to tune the binding affinity for cationic guests such as Na⁺ within the central O₆ cavity.

The di-exchanged complex, L³Co₂(pip)₂(OAc)₂, showed high affinity for cationic guests, owing to charge compensation of the dicationic L³Co₂ core by the newly introduced anionic acetato ligands. Indeed, Na⁺ ion was quantitatively taken up to form the inclusion complex L³Co₂(pip)₂(OAc)₂Na (Fig. 9g), whose structure was unambiguously determined by X-ray crystallography.

In contrast, when NaOTf was added to a CD₃CN solution of the starting piperidine-coordinating complex, L³Co₂(pip)₄, formation of the Na⁺ inclusion complex was not observed, indicating its lower guest binding affinity. However, upon treatment of this mixture with AcO⁻, ligand exchange rapidly proceeded, concomitant with Na⁺ uptake, to afford L³Co₂(pip)₂(OAc)₂Na (Fig. 9h). These observations clearly demonstrate that the guest binding affinity of the piperidine-coordinating complex L³Co₂(pip)₄ is enhanced by exchange of the neutral piperidine ligands with anionic acetato ligands.⁴³ Moreover, the presence of Na⁺ ion significantly enhanced the ligand exchange reactivity of the L³Co₂(pip)₄ metallohost.

Thus, the ligand exchange reactivity and the guest binding affinity of L³Co₂(pip)₄ were mutually enhanced. This interplay became even more evident in the guest binding of L³Co₂(pip)₄ accompanied by exchange with methoxo ligands under solvolytic conditions.

The piperidine-coordinating complex L³Co₂(pip)₄ slowly underwent solvolysis in CD₃OD, affording a new species rather than immediately forming a simple guest-bound complex, L³Co₂(pip)₄Na.⁴⁴ Spectroscopic and crystallographic investigations revealed that the resulting product was an inclusion complex L³Co₂(pip)₂(OMe)₂Na, in which two piperidine ligands were replaced by methoxo ligands. Detailed analysis of the reaction progress indicated that the reaction first produced a mono-exchanged guest-bound species, L³Co₂(pip)₃(OMe)Na, which was then converted into the di-exchanged species, L³Co₂(pip)₂(OMe)₂Na. In the first step, ligand exchange and Na⁺ uptake appear to occur concurrently.

More precisely, this process can be interpreted in terms of two possible mechanisms. One is the reaction first mechanism (Fig. 10A), in which exchange with the methoxo ligand occurs prior to guest binding. In this mechanism, the coordination of the anionic methoxo ligand cancels the positive charge of the cobalt(III) center, thereby enhancing the binding affinity for cationic guests in the central O₆ cavity. The other is the recognition first mechanism (Fig. 10B), in which guest binding occurs before ligand exchange. In this mechanism, the presence of the guest in the central O₆ cavity facilitates ligand exchange reaction at the cobalt(III) centers.

Fig. 10 Guest uptake behavior of L³Co₂(pip)₄ associated with the ligand exchange reaction under solvolytic conditions. The recognition first mechanism (B) is suggested when Na⁺ is bound.

These two mechanisms can be distinguished by analyzing the ligand exchange kinetics at varying guest concentrations.⁴⁵ In practice, the ligand exchange rate of L³Co₂(pip)₄ increased proportionally with the concentration of Na⁺, supporting the recognition first mechanism (Fig. 10B). In contrast, the ligand exchange rates observed in the presence of 1 equiv. of K⁺ or Rb⁺ were not significantly different from that without any guest, supporting the reaction first mechanism (Fig. 10A).⁴⁴ Thus, the mechanism of guest binding can be switched simply by changing the guest ions.

In enzymatic reactions, substrate binding is often accompanied by structural changes in the host protein. The well-known induced-fit mechanism describes a structural change in a protein that occurs as a result of substrate binding. More recently, an alternative mechanism known as conformational selection has been proposed, in which the conformational change precedes substrate binding.⁴⁶ The reaction first and recognition first mechanisms proposed in this study correspond to the conformational selection and induced-fit mechanisms, respectively, in enzymatic systems. Notably, this study provides the first demonstration that switching between these two mechanisms can be achieved simply by selecting different guest ions.

2.5. Ligand exchange of bis(saloph) cobalt(III) metallohosts with bridging diamines for guest recognition control

As described in the previous sections, L³Co₂(pip)₄ proved useful as the starting compound for the post-metalation modification to introduce various types of monodentate ligands, particularly primary amines. Its high reactivity toward primary amines also enabled the incorporation of diamine molecules bearing two terminal primary amino groups. Indeed, ligand exchange of L³Co₂(pip)₄ with 1,6-hexanediamine (DA1) afforded a doubly bridged metallohost, L³Co₂(DA1)₂, in 77% yield (Fig. 11a). This product has two bridging DA1 ligands on both faces of the macrocyclic plane.⁴⁷

Fig. 11 Formation of doubly and singly bridged metallohosts with the 1,6-hexanediamine ligand (DA1) based on the L³Co₂ macrocycle.

The utility of the ligand exchange strategy for synthesizing the doubly bridged complex L³Co₂(DA1)₂ was demonstrated by a failed attempt to obtain it directly from its components, H₄L³, Co(OAc)₂, and DA1, under aerobic conditions (Fig. 11b). This reaction instead yielded a singly bridged species, formulated as L³Co₂H(DA1)₂(OAc) (Fig. 11c). These results suggested that the singly bridged species is thermodynamically more favored in the presence of AcO⁻ than the doubly bridged species. Indeed, the doubly bridged species L³Co₂(DA1)₂ was gradually converted into the singly bridged species L³Co₂(DA1)₂(OAc) in the presence of AcO⁻, although this conversion was slow and incomplete (28% conversion after 24 h) (Fig. 11d).

Among the two types of bridged metallohosts, the doubly bridged species, L³Co₂(DA1)₂, showed a lower binding affinity for Na⁺ (K_a = 360 M⁻¹), which may be attributed to hindered access of counter anions due to the bridging ligands DA1. The Na⁺ binding affinity of the singly bridged species, L³Co₂H(DA1)₂(OAc), was only slightly higher (K_a = 670 M⁻¹), which can be rationalized by electrostatic repulsion from the protonated amino group (NH₃⁺) at the terminus of the non-bridging diamine ligand DA1. Indeed, once this complex was deprotonated with Bu₄NOAc (Fig. 11e), the resulting species, L³Co₂(DA1)₂(OAc), showed a significantly higher binding affinity for Na⁺ (K_a = 64 000 M⁻¹). Thus, the gate-opening from the doubly bridged to the singly bridged structure significantly enhanced guest binding affinity.⁴⁷

As described in the previous sections, the guest binding affinity and ligand exchange reactivity enhanced each other in the case of the piperidine-coordinating complex, L³Co₂(pip)₄. A similar mutual enhancement was observed between the gate-opening reactivity of the doubly bridged species, L³Co₂(DA1)₂, and the Na⁺ binding in the cavity. While the conversion of the doubly bridged species L³Co₂(DA1)₂ into the singly bridged species L³Co₂(DA1)₂(OAc) was slow (28% conversion after 24 h; k = 3.7 × 10⁻⁶ s⁻¹) as mentioned above (Fig. 11d), this reaction was significantly accelerated in the presence of Na⁺, by approximately 75-fold (k = 2.8 × 10⁻⁴ s⁻¹). Thus, the Na⁺ binding and the closed → open conversion of the doubly bridged species L³Co₂(DA1)₂ proceeded in a mutually promoted manner.⁴⁷

2.6. Shape-complementary introduction of bridging ligands into bis(saloph) cobalt(III) macrocycles

A larger class of macrocyclic bis(saloph) ligands was also employed in the synthesis of dinuclear structures, in which the two bridging ligands connect the cobalt(III) ions.⁴⁸ This doubly bridged complex was synthesized by the reaction of macrocycle H₄L⁴ (Fig. 2) with Co(OAc)₂ followed by the reaction with the bis-pyridine ligand BP1 (Fig. 12a). The resulting complex, L^4aCo₂(BP1)₂ (Fig. 12b), was characterized by NMR spectroscopy and mass spectrometry as well as X-ray crystallography. Given its structural similarity to Pd₂L₄ lantern-shaped cage structures,⁴⁹ the doubly bridged dinuclear cobalt(III) complex L^4aCo₂(BP1)₂ can be regarded as a trans-A₂B₂-type heteroleptic cage, because two types of arms alternately connect the two cobalt(III) ions. Previously, special strategies such as shape-complementary assembly were required for the synthesis of such heteroleptic cages.⁵⁰

Fig. 12 Lantern-shaped cage structures based on the bis(saloph) macrocycle H₄L⁴. (a) Bis(pyridine) ligands used as bridging ligands. (b) Symmetric doubly bridged complex L^4aCo₂(BP1)₂. (c) Unsymmetrical doubly bridged complexes L⁴Co₂(BP2)(BP3), which can be regarded as a lantern-shaped cage incorporating three or four different arms.

Whereas the L^4aCo₂(BP1)₂ complex has two identical bridging BP1 ligands on both faces, two different bridging ligands can also be selectively introduced into the L^4aCo₂ macrocycle via the shape-complementary assembly approach.⁵¹ Indeed, a heteroleptic doubly bridged complex, L^4aCo₂(BP2)(BP3), was successfully synthesized (Fig. 12c) using a similar protocol with a combination of the convergent-shaped BP2, which is based on carbazole, and the divergent-shaped BP3, which is based on fluorene (Fig. 12a).⁴⁸ X-ray crystallography clearly demonstrated the face-selective bridging of the two different bis-pyridine ligands, each geometrically suited to the bent L^4aCo₂ macrocyclic framework; the divergent-shaped BP3 binds to the concave face, while the convergent-shaped BP2 occupies the convex face. Notably, the corresponding reactions using only BP2 or BP3 with the L^4aCo₂ macrocycle failed to yield the doubly bridged structure. Detailed investigation revealed that the selective formation of the mixed-ligand species, L^4aCo₂(BP2)(BP3), is ascribed to its thermodynamic stability.

Furthermore, the unsymmetrical macrocycle H₄L^4b was employed in the synthesis of the mixed-ligand species L^4bCo₂(BP2)(BP3) (Fig. 12c). This complex has four different arms between the two cobalt(III) ions, which can be regarded as an M₂ABCD-type lantern-shaped cage, a structure that was previously considered difficult to synthesize.⁴⁸

3. Functioinalization of tris(saloph) cobalt(III) cages by axial coordination

3.1. Dynamic chirality inversion of helical tris(saloph) cobalt(III) cryptands by ligand exchange

A cryptand-like tris(saloph) cage ligand, H₆L⁵ (Fig. 2),⁵² features a bicyclic structure in which the three arms are doubly connected through two propeller-shaped triphenylbenzene subunits. Each arm contains a tetradentate H₂saloph chelate coordination site, which can accommodate a transition metal ion. The triply metalated species, L⁵M₃, is also expected to show strong binding affinity toward cationic guest species, due to the negatively polarized phenoxo groups in the [M(saloph)] substructures, as observed in the macrocyclic bis(saloph) analogues L³M₂. Indeed, the trinuclear nickel(II) complex L⁵Ni₃ (Fig. 13a) was found to exhibit unique binding behavior toward alkali metal ions, silver ion, guanidinium ion, etc.^52–54

Fig. 13 Chemical structures of the metallocryptands, (a) L⁵Ni₃ and (b) L⁵Co₃A₆.

This cryptand ligand H₆L⁵ can be converted into a trinuclear cobalt(III) complex, L⁵Co₃A₆. The six axial ligands A, coordinating to the three cobalt(III) ions in the saloph arms (Fig. 13b), can be replaced or modified, enabling functionalization and structural conversion of the cryptand framework. In fact, the trinuclear complexes L⁵Co₃A₆ (A = Me₂NH, pip) were synthesized by the reaction of H₆L⁵ with cobalt(II) acetate in the presence of appropriate amines A under aerobic conditions.⁵⁵

The three [Co(saloph)A₂]⁺ arms of the trinuclear complexes L⁵Co₃A₆ form a triple helix that extends toward the two propeller-shaped triphenylbenzene cores at the bridgeheads of the cryptand. In fact, the triple helical structure of the corresponding nickel(II) analogue, L⁵Ni₃, was confirmed by X-ray crystallography. This structure exhibits dynamic behavior, enabling interconversion between the P and M forms.^52,53 The present cobalt(III) analogues L⁵Co₃A₆, which contain achiral amine ligands (A = Me₂NH, pip), also adopt a similar helical structure. The P and M forms constitute an enantiomeric pair, yielding an equilibrated racemic mixture due to the dynamic chirality interconversion. However, ligand exchange of the [Co(saloph)A₂]⁺ (A = Me₂NH, pip) arms with a chiral amine ligand makes the P and M forms diastereomeric, thereby shifting the P/M equilibrium accordingly (Fig. 14a).

Fig. 14 Control of the P/M ratio of the helical metallocryptand L⁵Co₃A₆ via ligand exchange at the cobalt centers.

Indeed, the addition of chiral amines, S-A1 or S-A2, to L⁵Co₃A₆ (A = Me₂NH, pip) resulted in ligand exchange with these chiral amines, accompanied by a gradual P/M equilibrium shift (Fig. 14a).⁵⁵ The progress of ligand exchange at the cobalt(III) centers was easily monitored by ¹H NMR spectroscopy as well as mass spectrometry. For example, the dimethylamine-coordinating complex L⁵Co₃(Me₂NH)₆ was gradually converted to L⁵Co₃(S-A2)₆ upon addition of 12 equiv. of the chiral amine S-A2. This ligand exchange induced an equilibrium shift toward the P form, as evidenced by the growth of the negative CD signal at 550 nm (Fig. 15a(ii)). Time-course measurements indicated that the ligand exchange was almost completed within 3 h.

Fig. 15 Time-dependent changes in CD intensity at 550 nm observed during the ligand exchange of L⁵Co₃A₆. (a) CD intensity of L⁵Co₃A₆ (A = Me₂NH, pip) plotted versus time after the addition of S-A1 or S-A2 (12 equiv.). (b) CD intensity of L⁵Co₃(Me₂NH)₆ plotted versus time after sequential addition of S-A1 followed by a second chiral amine R-A1 or R-A2 (120 equiv.). Adapted with permission from ref. 55.

One notable advantage of this system is that the rate of the P/M equilibrium shift can be modulated by selecting different combinations of the initial achiral amine A and the added chiral amine (Fig. 15a(i–iv)). For example, when the chiral amine S-A1 was added instead of S-A2 to L⁵Co₃(Me₂NH)₆, the equilibrium shift proceeded more rapidly. In contrast, when the piperidine-coordinating complex L⁵Co₃(pip)₆ was used instead of L⁵Co₃(Me₂NH)₆ as the initial complex, the reaction became significantly slower. As a result, among the four possible combinations, the rate of the P/M equilibrium shift differed by up to 60-fold.⁵⁵

The ligand exchange strategy was also effective in inducing P/M chirality inversion when chiral amines with the opposite stereoconfigurations were used as the initial and added chiral sources (Fig. 14b). In fact, a P-major mixture was first prepared by the addition of the chiral amine S-A1 to racemic L⁵Co₃(Me₂NH)₆, which was then inverted to an M-favored mixture upon addition of a large excess of another chiral amine, R-A1, possessing the opposite stereoconfiguration. This P/M chirality inversion was monitored by time-dependent CD spectroscopy (Fig. 15b); upon addition of the second amine, the negative CD signal at 550 nm decreased and inverted within 8 min, then became nearly constant after 1 h. When another second chiral amine, R-A2, was used instead of R-A1, the P → M chirality inversion occurred 6 times more slowly. Thus, the L⁵Co₃A₆ complex was shown to be a useful helical molecular platform^56,57 capable of modulating the response speeds of both the P/M equilibrium shifts and the chirality inversion through the choice of initial and added amines.⁵⁵

The reverse ligand exchange, i.e., replacing a chiral ligand with an achiral ligand, showed a unique time-dependent change. When the achiral amine, piperidine, was added to L⁵Co₃(S-A1)₆, the six chiral amine ligands (S-A1) were gradually replaced by piperidine, yielding L⁵Co₃(pip)₆, which no longer contained any chiral sources in its molecular structure. This process converted a diastereomeric pair of L⁵Co₃(S-A1)₆ with a biased P/M ratio (P/M = 88 : 12) into an enantiomeric pair, which, in principle, could afford a racemic P/M mixture (Fig. 14c), assuming rapid P/M interconversion. Thus, the CD signal was expected to monotonically decrease and eventually become silent along the progress of the ligand exchange. However, the CD intensity exhibited an unusual irregular time-dependent change beyond expectations.⁵⁸

As already stated, L⁵Co₃(S-A1)₆ exhibited a negative CD signal at 550 nm due to its P-preference (Fig. 16a(i)), but the addition of a large excess of piperidine as an achiral amine (120 equiv.) caused an immediate decrease in the CD signal followed by a reversal instead of a simple monotonic decay to zero. The resulting positive signal reached its maximum intensity after 2 h (Fig. 16a(ii)), and then gradually diminished, eventually becoming CD-silent after 3 d (Fig. 16b(i, ii) and c). This behavior indicated that the predominant chirality transiently shifted from P to M before complete racemization occurred (Fig. 17).⁵⁸ Normally, when an optically active chiral compound undergoes racemization, its optical purity monotonically decreases to zero, and the chirality never inverts during the racemization process. Such a sign inversion before reaching equilibrium position is commonly observed in physical phenomena, such as damped oscillations, but is quite rare in the relaxation behavior of chemical reactions.

Fig. 16 Time-dependent changes in the CD spectra of metallocryptand L⁵Co₃(S-A1)₆ after addition of piperidine (120 equiv.). (a) Full CD spectra for 0–120 min and (b) after 120 min. (c) CD intensity at 550 nm plotted versus time. Adapted with permission from ref. 58.

Fig. 17 Overview of the transient P/M chirality inversion observed during the racemization of L⁵Co₃(S-A1)₆ via ligand exchange with piperidine.

This unusual and irregular time-course change during the racemization of L⁵Co₃(S-A1)₆ was investigated in detail by spectroscopic techniques, which clearly demonstrated that the six-step ligand exchange with piperidine was completed within 1 h to afford L⁵Co₃(pip)₆, i.e., at an early stage of the entire time-dependent process. The rate constants for all the ligand exchange steps were determined, revealing that the tetra-exchanged intermediate, L⁵Co₃(S-A1)₂(pip)₄, which still retained two chiral S-A1 ligands, preferred the reversed stereoconfiguration, namely the M helix (Fig. 17a). It then lost all the chiral S-A1 ligands while maintaining this M-biased configuration to afford L⁵Co₃(pip)₆ (Fig. 17b), which could undergo only slow racemization over 2–3 d (Fig. 17c) due to the steric hindrance imposed by the six bulky piperidine ligands.⁵⁸

Considering the above findings, opposite chiralities emerged during the forward (achiral → chiral; Fig. 14a) and reverse (chiral → achiral; Fig. 14c) ligand exchange reactions between piperidine and the chiral amine S-A1. During the ligand exchange of L⁵Co₃(pip)₆ with S-A1, the P form was always dominant (Fig. 14a), resulting in a monotonic increase in the P/M ratio. In contrast, the M form transiently became dominant during the ligand exchange of L⁵Co₃(S-A1)₆ with piperidine (Fig. 14c), leading to an unexpected transient chirality inversion (Fig. 17). This behavior constitutes a hysteretic cycle, because the M form appeared only in the reverse reaction. Such an unusual time-dependent change was discovered primarily because both the multi-step ligand exchange at the cobalt(III) centers and the P/M chirality inversion of the L⁵Co₃A₆ framework occurred on a similar timescale of minutes to hours, readily observable on a human time scale.

3.2. Tris(saloph) cobalt(III) closed cages for guest uptake/release control

Closed-cage type host molecules were anticipated to be formed when diamine ligands were introduced into the cryptand cage L⁵Co₃ in a bridging fashion between neighboring cobalt(III) ions at the cage apertures, as demonstrated in the macrocyclic dinuclear complex L³Co₂(DA1)₂. The resulting closed cage species, L⁵Co₃(diamine)₃ (Fig. 18), could serve as a molecular container, because guest uptake/release is completely blocked by the bridging diamine ligands, functioning much like a container with a cap or a lid. This effective gating arises from the kinetic inertness of the cobalt(III) centers, whose slow Co–N bond cleavage/formation even suppresses transient formation of open intermediates that could allow guest entry/exit. In contrast, the corresponding monoamine-coordinating analogue, L⁵Co₃A₆ (Fig. 13b), would behave as an open container, which allows rapid guest entry/exit. Thus, interconversion between the open and closed forms functions analogously to opening and closing a container lid,^59,60 providing control over the uptake and release of molecular and ionic guests.

Fig. 18 Chemical structures of closed metallocryptands, L⁵Co₃(diamine)₃.

The triply bridged closed metallohost, L⁵Co₃(DA1)₃, was synthesized by the reaction of the cryptand ligand H₆L⁵ with cobalt(II) acetate in the presence of diamine DA1 under aerobic conditions.⁶¹ This complex was also accessible via ligand exchange of the methylamine-coordinating open-cage complex, L⁵Co₃(MeNH₂)₆, with the diamine DA1. X-ray crystallography clearly revealed the triply bridged structure of this L⁵Co₃(DA1)₃, in which three DA1 molecules connect neighboring cobalt(III) ions to form a cyclic framework. These three diamine molecules are well accommodated in the grooves of the L⁵Co₃ triple helix, nearly completely sealing the apertures (Fig. 19a). Nevertheless, sufficient space remains at the center of the L⁵Co₃(DA1)₃ cryptand to accommodate a guest species.

Fig. 19 (a) X-ray crystal structure of L⁵Co₃(DA1)₃. (b) Guest uptake kinetics of L⁵Co₃(DA1)₃ in CD₃OD. Adapted with permission from ref. 61.

As expected, the bridging ligands effectively blocked the uptake of guest species such as alkali metal ions into the cavity of L⁵Co₃(DA1)₃, as evidenced by the ¹H NMR spectra recorded immediately after guest addition. Notably, larger alkali metal ions such as Cs⁺ and Rb⁺ were taken up only very slowly, requiring approximately 5 d to reach equilibrium (Fig. 19b). This indicates that the guest uptake was kinetically suppressed, rather than thermodynamically unfavorable, and that the closed cage L⁵Co₃(DA1)₃ retained its intrinsic affinity for guest binding. Kinetic analysis revealed that guest uptake into the closed cage L⁵Co₃(DA1)₃ proceeded at least 2000 times more slowly than into the corresponding open cage L⁵Co₃(MeNH₂)₆ (Fig. 20).⁶¹

Fig. 20 Slow uptake of Cs⁺ by closed metallocryptand L⁵Co₃(DA1)₃.

A related closed-cage metallocryptand, L⁵Co₃(DA7)₃, bearing cystamine bridging ligands DA7, was also synthesized (Fig. 18).⁶² This DA7 ligand contains a disulfide bond, which is known to behave dynamically in the presence of nucleophiles⁶³ and to remain static in their absence. Indeed, the presence of a thiolate anion was found to accelerate Cs⁺ uptake into L⁵Co₃(DA7)₃ by a factor of 14, whereas in its absence, the bridging DA7 ligands effectively blocked Cs⁺ uptake, as observed for the hexanediamine analogue, L⁵Co₃(DA1)₃.

The presence of the thiolate anion promoted the disulfide exchange equilibrium, transiently generating an open-cage intermediate that allows rapid guest entry/exit through the apertures (Fig. 21). In this system, the apertures of the L⁵Co₃ cage were initially sealed by the introduction of bridging ligands through coordination bond formation, thereby suppressing guest uptake/release. These apertures were then transiently opened in the presence of a nucleophile via disulfide exchange reactions.

Fig. 21 Proposed mechanism for guest uptake by L⁵Co₃(DA7)₃ via disulfide exchange accelerated by a thiolate ion.

4. Conclusions

Various types of cobalt(III) complexes bearing bis- and tris(saloph) ligands have been synthesized and shown to exhibit dynamic behaviors, despite the typically inert low-spin d⁶ electron configuration of cobalt(III) centers. This dynamic character arises primarily from the selective exchangeability of the two axial monodentate ligands in the [Co(saloph)X₂]⁺ units with ligands such as primary amines, pyridine, and anionic species. Such ligand exchange enables the functionalization and dynamic switching of multi-metallic structures, including reversible redox-driven dissociation/reassociation, modulation of host–guest binding affinity and kinetics, chirality control with time-dependent inversion, and gating of guest uptake through cage closure.

In some systems, guest binding also influences ligand exchange kinetics, expanding the functional scope of these coordination platforms. Thus, the controlled introduction, removal, and exchange of axial ligands has proven to be a powerful strategy for creating functional and responsive metallohosts and metallo-supramolecular systems.

At the same time, several challenges and limitations remain to be addressed. For example, the long-term stability of axial ligands under various conditions, the potential fatigue of the systems under repeated switching cycles, and limitations in guest selectivity are all important issues that require further investigation. Understanding and overcoming these challenges will be essential for translating these dynamic systems into practical stimuli-responsive materials and molecular devices.

Conflicts of interest

There are no conflicts to declare.

Data availability

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

Acknowledgements

The author warmly thanks all coworkers, collaborators, and past and present group members for their contributions to the work carried out in his laboratory. Special thanks are due to Prof. Yoko Sakata (Nagoya University), Dr Naoki Ousaka (Kyushu University), and Dr Sk Asif Ikbal (Kanazawa University) for their dedicated contributions, and to Prof. Tatsuya Nabeshima (University of Tsukuba) for valuable advice and discussions. The research cited in this article was supported by the Japan Society for the Promotion of Science (KAKENHI Grants JP21H05477 and JP23H04021 [Condensed Conjugation] as well as JP16H06510 [Coordination Asymmetry], JP18H03913, JP20K21206, JP23H01972, and JP23K17928), the CHOZEN Project at Kanazawa University, and the World Premier International Research Center Initiative (WPI), MEXT, Japan.

References

1. (a) P. Pfeiffer, E. Breith, E. Lübbe and T. Tsumaki, Liebigs Ann. Chem., 1933, 503, 84–130; (b) P. Pfeiffer, T. Hesse, H. Pfitzner, W. Scholl and H. Thielert, J. Prakt. Chem., 1937, 149, 217–296; (c) C. S. Marvel, S. A. Aspey and E. A. Dudley, J. Am. Chem. Soc., 1956, 78, 4905–4909.
2. (a) E. N. Jacobsen, Asymmetric Catalytic Epoxidation of Unfunctionalized Olefins in Catalytic Asymmetric Synthesis, ed. I. Ojima, VCH, New York, 1993, ch. 4.2, pp. 159–202; (b) T. Katsuki, Coord. Chem. Rev., 1995, 140, 189–214; (c) E. N. Jacobsen, Acc. Chem. Res., 2000, 33, 421–431.
3. N. Hoshino, Coord. Chem. Rev., 1998, 174, 77–108.
4. (a) G. Consiglio, I. P. Oliveri, S. Failla and S. di Bella, Molecules, 2019, 24, 2514; (b) S. di Bella, Dalton Trans., 2021, 50, 6050–6063.
5. H. Miyasaka, A. Saitoh and S. Abe, Coord. Chem. Rev., 2007, 251, 2622–2664.
6. (a) X. Yang, R. A. Jones and S. Huang, Coord. Chem. Rev., 2014, 273–274, 63–75; (b) S. di Bella and I. Fragalà, Synth. Met., 2000, 115, 191–196; (c) P. G. Lacroix, Eur. J. Inorg. Chem., 2001, 339–348.
7. (a) H.-Y. Yin, J. Tang and J.-L. Zhang, Eur. J. Inorg. Chem., 2017, 5085–5093; (b) A. Erxleben, Inorg. Chim. Acta, 2018, 472, 40–57; (c) L. A. Alfonso-Herrera, D. Hernández-Romero, J. A. Cruz-Navarro, Á. Ramos-Ligonio, A. López-Monteon, J. M. Rivera-Villanueva, D. Morales-Morales and R. Colorado-Peralta, Coord. Chem. Rev., 2024, 505, 215698.
8. (a) M. J. MacLachlan, Pure Appl. Chem., 2006, 78, 873–888; (b) A. W. Kleij, Chem. – Eur. J., 2008, 14, 10520–10529; (c) A. W. Kleij, Eur. J. Inorg. Chem., 2009, 193–205; (d) T. Glaser, Chem. Commun., 2011, 47, 116–130; (e) C. J. Whiteoak, G. Salassa and A. W. Kleij, Chem. Soc. Rev., 2012, 41, 622–631.
9. (a) T. Nabeshima and S. Akine, Chem. Rec., 2008, 8, 240–251; (b) S. Akine and T. Nabeshima, Dalton Trans., 2009, 10395–10408; (c) T. Nabeshima, S. Akine, C. Ikeda and M. Yamamura, Chem. Lett., 2010, 39, 10–16; (d) S. Akine, J. Inclusion Phenom. Macrocyclic Chem., 2012, 72, 25–54; (e) S. Akine, Metal Complexes with Oligo(salen)-Type Ligands in The Chemistry of Metal Phenolates, ed. J. Zabicky, Patai's Chemistry of Functional Groups, John Wiley & Sons Ltd, Chichester, 2018, ch. 4, vol. 2, pp. 153–194; (f) M. T. Chaudhry, S. Akine and M. J. MacLachlan, Chem. Soc. Rev., 2021, 50, 10713–10732.
10. (a) S. J. Wezenberg and A. W. Kleij, Angew. Chem., Int. Ed., 2008, 47, 2354–2364; (b) S. M. Elbert and M. Mastalerz, Org. Mater., 2020, 2, 182–203.
11. (a) R. M. Haak, S. J. Wezenberg and A. W. Kleij, Chem. Commun., 2010, 46, 2713–2723; (b) A. Decortes, A. M. Castilla and A. W. Kleij, Angew. Chem., Int. Ed., 2010, 49, 9822–9837; (c) D. E. White, P. M. Tadross, Z. Lu and E. N. Jacobsen, Tetrahedron, 2014, 70, 4165–4180.
12. For acyclic salen/saloph oligomers, see: (a) J. H. Chong, S. J. Ardakani, K. J. Smith and M. J. MacLachlan, Chem. – Eur. J., 2009, 15, 11824–11828; (b) S. J. Wezenberg, G. Salassa, E. C. Escudero-Adán, J. Benet-Buchholz and A. W. Kleij, Angew. Chem., Int. Ed., 2011, 50, 713–716; (c) T. J. Dunn, C. F. Ramogida, C. Simmonds, A. Paterson, E. W. Y. Wong, L. Chiang, Y. Shimazaki and T. Storr, Inorg. Chem., 2011, 50, 6746–6755; (d) M. Mastalerz and I. M. Oppel, Eur. J. Org. Chem., 2011, 5971–5980; (e) T. J. Dunn, L. Chiang, C. F. Ramogida, M. I. Webb, D. Savard, M. Sakaguchi, T. Ogura, Y. Shimazaki and T. Storr, Dalton Trans., 2012, 41, 7905–7914; (f) H. Achira, M. Ito, T. Mutai, I. Yoshikawa, K. Araki and H. Houjou, Dalton Trans., 2014, 43, 5899–5907; (g) S. A. Ikbal, Y. Sakata and S. Akine, Dalton Trans., 2021, 50, 4119–4123.
13. (a) M. E. Briggs and A. I. Cooper, Chem. Mater., 2017, 29, 149–157; (b) M. Mastalerz, Acc. Chem. Res., 2018, 51, 2411–2422.
14. (a) S. J. Rowan, S. J. Cantrill, G. R. L. Cousins, J. K. M. Sanders and J. F. Stoddart, Angew. Chem., Int. Ed., 2002, 41, 898–952; (b) M. Mastalerz, Angew. Chem., Int. Ed., 2010, 49, 5042–5053; (c) Y. Jin, C. Yu, R. J. Denman and W. Zhang, Chem. Soc. Rev., 2013, 42, 6634–6654.
15. (a) A. G. Manfredotti and C. Guastini, Acta Crystallogr., Sect. C: Cryst. Struct. Commun., 1983, C39, 863–865; (b) A. Radha, M. Seshasayee, K. Ramalingam and G. Aravamudan, Acta Crystallogr., Sect. C: Cryst. Struct. Commun., 1985, 41, 1169–1171; (c) J. Wang, F.-L. Bei, X.-Y. Xu, X.-J. Yang and X. Wang, J. Chem. Crystallogr., 2003, 33, 845–849.
16. (a) G. D. Fallon and B. M. Gatehouse, Acta Crystallogr., Sect. B, 1976, 32, 2591–2597; (b) N. Kumari, R. Prajapati and L. Mishra, Polyhedron, 2008, 27, 241–248; (c) S. Mahato, P. Rawal, A. K. Devadkar, M. Joshi, A. R. Choudhury, B. Biswas, P. Gupta and T. K. Panda, Org. Biomol. Chem., 2022, 20, 1103–1111; (d) Y. Sekine, S. Kusumoto, A. Sugimoto, M. Nakaya and S. Hayami, Cryst. Growth Des., 2023, 23, 2013–2017.
17. (a) D. Hall and F. H. Moore, J. Chem. Soc. A, 1966, 1822–1824; (b) C. A. Barboza, J. C. Germino, A. M. Santana, F. J. Quites, P. A. M. Vazquez and T. D. Z. Atvars, J. Phys. Chem. C, 2015, 119, 6152–6163.
18. (a) J. Reglinski, S. Morris and D. E. Stevenson, Polyhedron, 2002, 21, 2175–2182; (b) J. S. Matalobos, A. M. García-Deibe, M. Fondo, D. Navarro and M. R. Bermejo, Inorg. Chem. Commun., 2004, 7, 311–314; (c) A. W. Kleij, M. Kuil, M. Lutz, D. M. Tooke, A. L. Spek, P. C. J. Kamer, P. W. N. M. van Leeuwen and J. N. H. Reek, Inorg. Chim. Acta, 2006, 359, 1807–1814; (d) M. Odoko, N. Tsuchida and N. Okabe, Acta Crystallogr., Sect. E: Struct. Rep. Online, 2006, E62, m708–m709.
19. For a review, see: A. W. Kleij, Dalton Trans., 2009, 4635–4639.
20. (a) A. W. Kleij, M. Kuil, D. M. Tooke, M. Lutz, A. L. Spek and J. N. H. Reek, Chem. – Eur. J., 2005, 11, 4743–4750; (b) M. Kuil, P. E. Goudriaan, P. W. N. M. van Leeuwen and J. N. H. Reek, Chem. Commun., 2006, 4679–4681; (c) M. Kuil, I. M. Puijk, A. W. Kleij, D. M. Tooke, A. L. Spek and J. N. H. Reek, Chem. – Asian J., 2009, 4, 50–57; (d) M. Kuil, P. E. Goudriaan, A. W. Kleij, D. M. Tooke, A. L. Spek, P. W. N. M. van Leeuwen and J. N. H. Reek, Dalton Trans., 2007, 2311–2320; (e) T. Gadzikwa, R. Bellini, H. L. Dekker and J. N. H. Reek, J. Am. Chem. Soc., 2012, 134, 2860–2863.
21. (a) A. W. Kleij, M. Kuil, D. M. Tooke, A. L. Spek and J. N. H. Reek, Inorg. Chem., 2007, 46, 5829–5831; (b) J. Flapper and J. N. H. Reek, Angew. Chem., Int. Ed., 2007, 46, 8590–8592; (c) G. Salassa, A. M. Castilla and A. W. Kleij, Dalton Trans., 2011, 40, 5236–5243; (d) I. Jacobs, A. C. T. van Duin, A. W. Kleij, M. Kuil, D. M. Tooke, A. L. Spek and J. N. H. Reek, Catal. Sci. Technol., 2013, 3, 1955–1963; (e) D. Anselmo, R. Gramage-Doria, T. Besset, M. V. Escárcega-Bobadilla, G. Salassa, E. C. Escudero-Adán, M. M. Belmonte, E. Martin, J. N. H. Reek and A. W. Kleij, Dalton Trans., 2013, 42, 7595–7603.
22. M. Karmakar and S. Chattopadhyay, J. Mol. Struct., 2019, 1186, 155–186.
23. (a) M. Amirnasr, K. J. Schenk, A. Gorji and R. Vafazadeh, Polyhedron, 2001, 20, 695–702; (b) M. Salehi, S. Meghdadi, M. Amirnasr and K. Mereiter, Acta Crystallogr., Sect. E:Struct. Rep. Online, 2009, 65, m942–m943.
24. Ruthenium complex was also used as a platform of functionalization, see: K. Chichak, U. Jacquemard and N. R. Branda, Eur. J. Inorg. Chem., 2002, 357–368.
25. (a) S. C. F. Au-Yeung and D. R. Eaton, Inorg. Chem., 1984, 23, 1517–1520; (b) M. F. Summers, L. G. Marzilli, N. Bresciani-Pahor and L. Randaccio, J. Am. Chem. Soc., 1984, 106, 4478–4485; (c) L. M. Manus, R. J. Holbrook, T. A. Atesin, M. C. Heffern, A. S. Harney, A. L. Eckermann and T. J. Meade, Inorg. Chem., 2013, 52, 1069–1076.
26. S. Akine, Dalton Trans., 2021, 50, 4429–4444.
27. (a) E. Eichhorn, A. Rieker, B. Speiser and H. Stahl, Inorg. Chem., 1997, 36, 3307–3317; (b) A. A. Khandar, B. Shaabani, F. Belaj and A. Bakhtiari, Inorg. Chim. Acta, 2007, 360, 3255–3264.
28. (a) M. J. Hardie and C. L. Raston, Angew. Chem., Int. Ed., 2000, 39, 3835–3839; (b) D. A. Leigh, P. J. Lusby, R. T. McBurney, A. Morelli, A. M. Z. Slawin, A. R. Thomson and D. B. Walker, J. Am. Chem. Soc., 2009, 131, 3762–3771; (c) T. Nishino, Y. Yamada, S. Akine, K. Sugimoto and K. Tanaka, Dalton Trans., 2016, 45, 3831–3837; (d) Y.-L. Lu, K. Wu, Y.-H. Huang, W.-C. Li, Z.-M. Cao, X.-H. Yan, X.-D. Zhang, C.-H. Liu, J. Ruan, H.-S. Xu, M. Pan and C.-Y. Su, J. Am. Chem. Soc., 2024, 146, 20414–20424.
29. H. Shimakoshi, H. Takemoto, I. Aritome and Y. Hisaeda, Tetrahedron Lett., 2002, 43, 4809–4812.
30. H. Shimakoshi, T. Takemoto, I. Aritome and Y. Hisaeda, Inorg. Chem., 2005, 44, 9134–9136.
31. H. Shimakoshi, S. Hirose, M. Ohba, T. Shiga, H. Ōkawa and Y. Hisaeda, Bull. Chem. Soc. Jpn., 2005, 78, 1040–1046.
32. H. Shimakoshi, K. Shibata and Y. Hisaeda, Inorg. Chem., 2009, 48, 1045–1052.
33. H. Shimakoshi and Y. Hisaeda, J. Synth. Org. Chem., Jpn., 2012, 70, 60–70.
34. (a) S. Akine, F. Utsuno and T. Nabeshima, Chem. Commun., 2010, 46, 1029–1031; (b) S. Akine, F. Utsuno, S. Piao, H. Orita, S. Tsuzuki and T. Nabeshima, Inorg. Chem., 2016, 55, 810–821.
35. (a) M. Cametti, Y. Sakata, J. Martí-Rujas and S. Akine, Inorg. Chem., 2019, 58, 14871–14875; (b) M. T. Chaudhry, B. O. Patrick, S. Akine and M. J. MacLachlan, Org. Biomol. Chem., 2022, 20, 8259–8268; (c) S. Akine, M. Nakano, Y. Sakata and S. Yano, Chem. – Eur. J., 2024, 30, e202403071.
36. For related macrocyclic bis(saloph) hosts with different sizes, see: (a) F. C. J. M. van Veggel, M. Bos, S. Harkema, W. Verboom and D. N. Reinhoudt, Angew. Chem., Int. Ed. Engl., 1989, 28, 746–748; (b) F. C. J. M. van Veggel, M. Bos, S. Harkema, H. van de Bovenkamp, W. Verboom, J. Reedijk and D. N. Reinhoudt, J. Org. Chem., 1991, 56, 225–235; (c) Y. Sakata, S. Kobayashi and S. Akine, Chem. Commun., 2017, 53, 6363–6366; (d) Y. Sakata, S. Kobayashi, M. Yamamoto, K. Doken, M. Kamezawa, S. Yamaki and S. Akine, Commun. Chem., 2024, 7, 166; (e) S. Akine, Z. Varadi and T. Nabeshima, Eur. J. Inorg. Chem., 2013, 5987–5998.
37. For representative macrocyclic tris(saloph) hosts, see: (a) S. Akine, T. Taniguchi and T. Nabeshima, Tetrahedron Lett., 2001, 42, 8861–8864; (b) A. J. Gallant and M. J. MacLachlan, Angew. Chem., Int. Ed., 2003, 42, 5307–5310; (c) S. Akine, D. Hashimoto, T. Saiki and T. Nabeshima, Tetrahedron Lett., 2004, 45, 4225–4227; (d) T. Nabeshima, H. Miyazaki, A. Iwasaki, S. Akine, T. Saiki, C. Ikeda and S. Sato, Chem. Lett., 2006, 35, 1070–1071; (e) S. Akine, S. Sunaga, T. Taniguchi, H. Miyazaki and T. Nabeshima, Inorg. Chem., 2007, 46, 2959–2961; (f) A. Yamashita, A. Watanabe, S. Akine, T. Nabeshima, M. Nakano, T. Yamamura and T. Kajiwara, Angew. Chem., Int. Ed., 2011, 50, 4016–4019; (g) S. Akine, S. Sunaga and T. Nabeshima, Chem. – Eur. J., 2011, 17, 6853–6861; (h) T. Nakamura, S. Yano, Y.-F. Liu and T. Nabeshima, Inorg. Chem., 2024, 63, 12697–12702.
38. For related macrocyclic tetrakis(saloph) hosts, see: (a) S.-S. Sun, C. L. Stern, S. T. Nguyen and J. T. Hupp, J. Am. Chem. Soc., 2004, 126, 6314–6326; (b) S.-i. Kawano, Y. Ishida and K. Tanaka, J. Am. Chem. Soc., 2015, 137, 2295–2302; (c) T. Nakamura, S. Tsukuda and T. Nabeshima, J. Am. Chem. Soc., 2019, 141, 6462–6467; (d) S.-i. Kawano, M. Nakaya, M. Saitow, A. Ishiguro, T. Yanai, J. Onoe and K. Tanaka, J. Am. Chem. Soc., 2022, 144, 6749–6758.
39. Y. Sakata, C. Murata and S. Akine, Nat. Commun., 2017, 8, 16005.
40. (a) C. Marquez and W. M. Nau, Angew. Chem., Int. Ed., 2001, 40, 3155–3160; (b) O. Danylyuk, V. P. Fedin and V. Sashuk, Chem. Commun., 2013, 49, 1859–1861.
41. D. Walter, Y. Sakata and S. Akine, Chem. – Asian J., 2025, 20, e202401876.
42. (a) D. A. Roberts, B. S. Pilgrim and J. R. Nitschke, Chem. Soc. Rev., 2018, 47, 626–644; (b) M.-M. Gan, J.-Q. Liu, L. Zhang, Y.-Y. Wang, F. E. Hahn and Y.-F. Han, Chem. Rev., 2018, 118, 9587–9641.
43. Y. Sakata, M. Okada, M. Tamiya and S. Akine, Chem. – Eur. J., 2020, 26, 7595–7601.
44. Y. Sakata, M. Tamiya, M. Okada and S. Akine, J. Am. Chem. Soc., 2019, 141, 15597–15604.
45. (a) P. J. Tummino and R. A. Copeland, Biochemistry, 2008, 47, 5481–5492; (b) G. G. Hammes, Y.-C. Chang and T. G. Oas, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 13737–13741; (c) A. D. Vogt and E. di Cera, Biochemistry, 2012, 51, 5894–5902.
46. (a) L. C. James and D. S. Tawfik, Trends Biochem. Sci., 2003, 28, 361–368; (b) D. D. Boehr, R. Nussinov and P. E. Wright, Nat. Chem. Biol., 2009, 5, 789–796; (c) P. Csermely, R. Palotai and R. Nussinov, Trends Biochem. Sci., 2010, 35, 539–546.
47. Y. Sakata, M. Okada and S. Akine, Chem. – Eur. J., 2021, 27, 2284–2288.
48. B. Zhang, H. Lee, J. J. Holstein and G. H. Clever, Angew. Chem., Int. Ed., 2024, 63, e202404682.
49. D. A. McMorran and P. J. Steel, Angew. Chem., Int. Ed., 1998, 37, 3295–3297.
50. (a) W. M. Bloch and G. H. Clever, Chem. Commun., 2017, 53, 8506–8516; (b) S. Pullen, J. Tessarolo and G. H. Clever, Chem. Sci., 2021, 12, 7269–7293.
51. (a) W. M. Bloch, Y. Abe, J. J. Holstein, C. M. Wandtke, B. Dittrich and G. H. Clever, J. Am. Chem. Soc., 2016, 138, 13750–13755; (b) W. M. Bloch, J. J. Holstein, W. Hiller and G. H. Clever, Angew. Chem., Int. Ed., 2017, 56, 8285–8289.
52. (a) S. Akine, M. Miyashita, S. Piao and T. Nabeshima, Inorg. Chem. Front., 2014, 1, 53–57; (b) S. Akine, M. Miyashita and T. Nabeshima, Inorg. Chem., 2021, 60, 12961–12971.
53. S. A. Ikbal, P. Zhao, M. Ehara and S. Akine, Sci. Adv., 2023, 9, eadj5536.
54. For related oligo(salen)/oligo(saloph) cages, see: (a) S. Akine, S. Piao, M. Miyashita and T. Nabeshima, Tetrahedron Lett., 2013, 54, 6541–6544; (b) S. M. Elbert, F. Rominger and M. Mastalerz, Chem. – Eur. J., 2014, 20, 16707–16720; (c) J. Jiao, C. Tan, Z. Li, Y. Liu, X. Han and Y. Cui, J. Am. Chem. Soc., 2018, 140, 2251–2259; (d) S. M. Elbert, W.-S. Zhang, Y. Vaynzof, N. Oberhof, M. Bernhardt, M. Pernpointner, F. Rominger, R. R. Schröder and M. Mastalerz, Chem. Mater., 2019, 31, 6210–6223; (e) Y. Sakata, R. Tsuyuki, S. Sugimoto and S. Akine, Chem. Commun., 2021, 57, 13510–13513; (f) F. Uhrmacher, S. M. Elbert, F. Rominger and M. Mastalerz, Eur. J. Inorg. Chem., 2022, e202100864; (g) S. Firdausiah, R. Ide, M. Ehara, P. Zhao, Y. Sakata and S. Akine, Chem. – Eur. J., 2025, 31, e202501277.
55. Y. Sakata, S. Chiba, M. Miyashita, T. Nabeshima and S. Akine, Chem. – Eur. J., 2019, 25, 2962–2966.
56. (a) H. Miyake and H. Tsukube, Chem. Soc. Rev., 2012, 41, 6977–6991; (b) S. Akine, Inorganics, 2018, 6, 80; (c) S. Akine and H. Miyake, Coord. Chem. Rev., 2022, 468, 214582; (d) S. Akine, Pure Appl. Chem., 2023, 95, 331–341.
57. Recent examples of chirality control of metal complexes, see: (a) S. Akine, T. Taniguchi and T. Nabeshima, Tetrahedron Lett., 2006, 47, 8419–8422; (b) S. Akine, T. Matsumoto and T. Nabeshima, Chem. Commun., 2008, 4604–4606; (c) S. Akine, S. Hotate, T. Matsumoto and T. Nabeshima, Chem. Commun., 2011, 47, 2925–2927; (d) S. Akine, S. Hotate and T. Nabeshima, J. Am. Chem. Soc., 2011, 133, 13868–13871; (e) S. Akine, S. Sairenji, T. Taniguchi and T. Nabeshima, J. Am. Chem. Soc., 2013, 135, 12948–12951; (f) S. Sairenji, S. Akine and T. Nabeshima, Tetrahedron Lett., 2014, 55, 1987–1990; (g) S. Sairenji, S. Akine and T. Nabeshima, Chem. Lett., 2014, 43, 1107–1109; (h) S. Akine, T. Matsumoto and T. Nabeshima, Angew. Chem., Int. Ed., 2016, 55, 960–964; (i) S. Sairenji, S. Akine and T. Nabeshima, Dalton Trans., 2016, 45, 14902–14906; (j) S. Sairenji, S. Akine and T. Nabeshima, Sci. Rep., 2018, 8, 137; (k) S. Akine, K. Nomura, M. Takahashi, Y. Sakata, T. Mori, W. Nakanishi and K. Ariga, Dalton Trans., 2023, 52, 260–268; (l) S. Muratsugu, K. Sawaguchi, T. Shiraogawa, S. Chiba, Y. Sakata, S. Shirai, H. Baba, M. Ehara, S. Akine and M. Tada, Chem. Commun., 2024, 60, 2094–2097.
58. Y. Sakata, S. Chiba and S. Akine, Proc. Natl. Acad. Sci. U. S. A., 2022, 119, e2113237119.
59. S. Akine and Y. Sakata, Chem. Lett., 2020, 49, 428–441.
60. (a) E. L. Piatnitski and K. D. Deshayes, Angew. Chem., Int. Ed., 1998, 37, 970–972; (b) S. Zarra, M. M. J. Smulders, Q. Lefebvre, J. K. Clegg and J. R. Nitschke, Angew. Chem., Int. Ed., 2012, 51, 6882–6885; (c) H. Wang, F. Liu, R. C. Helgeson and K. N. Houk, Angew. Chem., Int. Ed., 2013, 52, 655–659.
61. S. Akine, M. Miyashita and T. Nabeshima, J. Am. Chem. Soc., 2017, 139, 4631–4634.
62. S. Akine, M. Miyashita and T. Nabeshima, Chem. – Eur. J., 2019, 25, 1432–1435.
63. S. P. Black, J. K. M. Sanders and A. R. Stefankiewicz, Chem. Soc. Rev., 2014, 43, 1861–1872.

---

Footnotes

† H₂salen = N,N′-disalicylideneethylenediamine.

‡ H₂saloph = N,N′-disalicylidene-o-phenylenediamine.

---