Bimetallic Chemistry by Design: Approaches to the Development of Multifunctional Ligands

Abstract

Bimetallic complexes enable cooperative function that is inaccessible at mononuclear sites, yet realizing such behavior in a predictable manner requires precise control over metal identity, metal/metal separation and bridging ligand environments. This review surveys synthetic strategies for constructing bimetallic architectures and highlights how these concepts inform the rational design of multifunctional ligand platforms. We focus on the development of a tripodal ligand with phosphinic amido groups [poat]³⁻ which enables the site-directed assembly of heterobimetallic cores. This framework provides access to FeFe, FeMn and CoFe complexes spanning multiple oxidation and protonation states, allowing systematic interrogation of proton-coupled electron transfer, redox thermodynamics, and magnetic exchange. Comparative studies reveal how metal substitution modulates oxidizing power, PCET reactivity, and spin-state behavior, offering molecular-level insight into cooperative principles that underlie both metalloenzyme function and functional bimetallic materials.

Graphical Abstract

1. Introduction

1.1. Utility of bimetallic systems in chemistry

The development of molecular systems containing two metal ions has been a central theme in coordination chemistry for several decades. The appeal of bimetallic systems lies in their ability to exploit metal-metal cooperativity to accomplish a variety of functions not otherwise possible with a single metal site. By positioning two metal centers in proximity, bimetallic architectures can support multielectron redox chemistry, cooperative substrate activation, and magnetic exchange coupling. As a result, bimetallic motifs are pervasive across all areas of inorganic chemistry, ranging from functional materials to the active sites of metalloenzymes.

Establishing rational design principles that enable precise control over the positioning, electronic communication, and functional differentiation of two metal centers within a single molecular framework remains a major challenge in inorganic chemistry. Ligand design has emerged as a powerful tool to tackle this challenge, enabling chemists to dictate metal-metal separation, coordination geometry, electronic communication, and site selectivity, and to ultimately isolate many different types of homo- and hetero- bimetallic systems. This review will describe some of these ligand design approaches for the synthesis of bimetallic complexes, with particular emphasis on some of our contributions towards designing multifunctional ligands that allow for site-directed stepwise assembly of molecular bimetallic species.

1.2. Synthetic approaches: binucleating ligands

Among the earliest and most influential strategies for constructing bimetallic complexes is the use of binucleating ligands, most notably those pioneered by Robson [1]. Robson introduced the concept of ligands deliberately preorganized to secure two metal centers in proximity, with the explicit goal of enabling new types of reactivity, metal-metal electronic communication, and parallels to cofactors within metalloprotein active sites. These early designs demonstrated that enforcing short distances between metal centers could give rise to unprecedented structures and reaction pathways that were the direct consequences of ligand preorganization [2–5]. Robson-type ligands generate frameworks containing two well-defined metal-binding sites, imposing a fixed spacial separation typically 3–5 Å between metal centers, while offering a tunable array of donor groups to modulate coordination geometry and electronic structure (Fig. 1A). The isolation of the first macrocyclic binucleating ligand complex in the late 1960s, followed by a family of binuclear complexes spanning multiple transition metal ions, established a powerful new paradigm in coordination chemistry (Fig. 1B) [1]. Note that Robson’s conceptual emphasis on preorganization and ligand-directed function has presaged many of the design principles that later became central to metal-organic frameworks as illustrated by his share of the 2025 Nobel Prize. Since then, such ligands have been widely adopted as versatile platforms for probing multielectron redox chemistry, modeling enzymatic active sites, and constructing catalysts for small-molecule activation.

Figure 1.

Robson’s approach to designing systems capable of securing two metal centers in very close proximity, termed binucleating ligands, that offered new types of reactivity, catalysts, metal-metal electronic interaction, and useful parallels with metalloproteins (A). Robson’s first isolated crystalline complex of a binucleating ligand (B).

Representative examples that utilized these ideas come from the work of Sorrell who leveraged phenolate-bridged frameworks to prepare Cu/Cu complexes and monitor reactions with dioxygen [6]. Karlin used a similar binucleating ligand with an aryl spacer and found that the ring was hydroxylated when the Cu^I/Cu^I species reacted with dioxygen [7]. Que likewise prepared a series of phenolate-bridged heterobimetallic Fe/M (Mⁿ = Mn, Fe, Co, Ni, Zn; n= II or III) complexes, enabling systematic interrogation of individual metal ion effects on magnetic coupling and electronic structure (Fig. 2A) [8,9]. Subsequent work by Que, Karlin, Suzuki, and Kodera elaborated on these systems to form stable dioxygen adducts between two metal centers (Fig. 2B) [10–13]. Meyer also utilized binucleating systems featuring a pyrazolate bridge, allowing modular control over metal-metal separation and enabling detailed studies of small-molecule binding and activation (Fig. 2C) [14–17]. More recently, Buss has expanded this platform to access a broader range of heterobimetallic compositions (Fig. 2A) [18]. Parallel efforts by Zhang (Fig. 2D) and Tilley (Fig. 2E) introduced alternative binucleating ligands that provided insight into the metalation process relevant to metallocofactor assembly in proteins [19,20].

Figure 2.

Representative examples of binucleating ligands prepared by Que (A), Que, Suzuki and Kodera to control dioxygen binding (B), Meyer (C), Zhang (D), and Tilley (E).

1.3. Synthetic approaches: self-assembly

Another successful strategy for accessing bimetallic complexes relies on self-assembly, wherein discrete dinuclear species form as thermodynamic products. This approach has been particularly effective for modelling biological metallocofactors, especially, diiron and dimanganese cores involved in dioxygen binding and activation. Seminal work by Wieghardt and Lippard have both shown that using tridentate ligands such as 1,4,7-triazacyclononane (TACN) or trispyrazolylborates in the presence of bridging acetate ligands, complexes self-assemble to form triply bridged dinuclear cores (M^III–(μ-O)(μ−1,3-OAc)₂–M^III) (Fig. 3A,B) [21–26]. These motifs closely resemble the active site of the dioxygen-binding protein hemerythrin. Wieghardt further showed that modifying the substituents on the TACN ligand allowed fine control over the assembly process, leading to a family of bimetallic complexes with distinct primary coordination spheres [27–29]. Kitajima reported similar steric control with pyrazolate-based systems, where control of the steric bulk on the pyrazole rings directed the formation of alternative dinuclear structures. Notably, introduction of isopropyl substituents at the 3- and 5- positions of the pyrazole rings enabled isolation of a Cu/Cu complex that bound dioxygen in a μ-η²-η² fashion, preceding later structural elucidation of dioxygen binding in hemocyanin (Fig. 3C) [30]. Collectively, these studies underscore the power of self-assembly routes to biologically relevant bimetallic architectures.

Figure 3.

Representative examples of bimetallic complexes prepared using a self-assembly strategy by Wieghardt (A), Lippard (B), and Kitajima (C).

1.4. Synthetic approaches: multifunctional ligands (metalloligands)

More recent strategies for constructing bimetallic complexes have focused on multifunctional ligands, in which a pre-formed metal complex serves as a synthon (or metalloligand) for binding a second metal-containing fragment. A prominent example of this strategy was developed by Thomas, who pioneered the use of [Zr^IV(NR′PR₂)₃X] ligands, which can readily coordinate an additional metal center, most notably Co ions, through the phosphine groups to form a [Zr^IV(NR′PR₂)₃Coⁿ] (n = −1, 0, +1) (Fig. 4A) [31–33]. The resulting complexes exhibit C₃ symmetry, positioning two metal centers <2.5 Å apart, indicating the presence of metal/metal interactions. This enforced proximity has enabled access to rich and unusual reactivity patterns that arise from cooperative electronic effects between the two metals. Related tripodal metalloligand platforms have been developed by Lu (Fig. 4B) and Takaya/Iwasawa (Fig.4C), which function in an analogous manner by pre-organizing donor groups around a primary metal center to recruit a secondary metal ion [34–38]. These metalloligand-based strategies illustrate how preformed metal complexes can function as modular synthons for assembling heterobimetallic complexes, a design principle that informs on the site-directed approaches described in the following section.

Figure 4.

Multifunctional ligands for the preparation of heterobimetallic complexes prepared by Thomas (A), Lu (B), and Takaya and Iwasawa (C).

1.5. Molecular designs for the convergent preparation of bimetallic complexes

Our group has focused efforts on designing ligand scaffolds that emulate aspects of metalloenzymes to understand the effects of the local environment on the functions of metal centers. We have designed rigid ligand scaffolds that incorporate strategically positioned intramolecular hydrogen bond (H-bond) donors and acceptors that stabilize biologically relevant reactive intermediates and modulate reactivity [39–41]. Key ligand design features include: a trianionic metal binding pocket to support high valent metal centers, an N4 primary coordination environment, local C₃ symmetry about the metal ion, and the possibility to enforce up to three intramolecular H-bonds (Fig. 5). One such ligand is a urea-based tripod [H₃buea]³⁻ which supports the formation of [Fe^III/IVH₃buea(O)]^2−/− complexes (Fig. 6A); however, the steric bulk of the ethylene backbone and the tBu groups appended from the ureas prevents the assembly of bimetallic species using this ligand framework [39,42–47] (Table 1).

Figure 5.

Ligand design features to confine metal centers and stabilize high valent species.

Figure 6.

Fe complexes supported by tripodal ligand scaffolds designed in the Borovik laboratory. An Fe^IV=O species supported by the [H₃buea]³⁻ ligand, which has three urea arms that serve as H-bond donors (A). An Fe^II(OH₂) species supported by the [MST]³⁻ ligand, which contains H-bond accepting sulfonamido groups (B). An Fe^IV=O species supported by the [poat]³⁻ ligand (C).

Table 1.

Development of ligand platforms to support (hetero)bimetallic assembly.

Ligand Platform	H-bonding groups	Bimetallic Assembly	Accessible Oxidation States	Accessible Oxidation States for Auxiliary Metal
[H3buea]3-	Urea (H-bond donors)	No	2+, 3+, 4+	--
[MST]3-	Sulfonamido (H-bond acceptors)	Yes	2+, 3+	2+
[poat]3-	Phosphinic amido (H-bond acceptors)	Yes	2+, 3+, 4+	2+, 3+

We circumvent the steric crowding by designing the [MST]³⁻ ligand to maintain a similar primary coordination environment and symmetry of [H₃buea]³⁻ while introducing H-bond accepting sulfonamido groups (Fig. 6B and Table 1). An unexpected result of including sulfonamido moieties was that two of the S=O groups, along with the hydroxido ligand, created an auxiliary binding site for a second metal complex that allowed for controlled preparation of bimetallic complexes (Scheme 1) [45,48–52]. The [MST]³⁻ ligands allowed for the preparation of bimetallic complexes in which the metal ions have different primary coordination spheres; however, there were some limitations with this design approach. Notably, the anionic N-donors of the sulfonamido groups do not provide a sufficiently strong ligand field to support a metal center beyond the 3+ oxidation state. In addition, we found that it was difficult to oxidize the auxiliary metal center beyond the 2+ oxidation state, precluding preparation of high valent bimetallic complexes, especially those implicated as intermediates in biological systems [53].

Scheme 1.

General preparation of heterobimetallic complexes using the [MST]³⁻ ligand.

2. Development of Phosphinic Amido Ligands

We combined the advantageous properties of [H₃buea]³⁻ and [MST]³⁻ while addressing limitations in preparing highly oxidized bimetallic complexes by redesigning a new tripodal ligand, [poat]³⁻ that is based on phosphinic amido groups [54–58] (Table 1). We have previously reviewed the coordination chemistry with the [H₃buea]³⁻ and [MST]³⁻ ligands [40,41,47]; this review will cover recent developments on chemistry involving the [poat]³⁻ ligand. The incorporation of phosphinic amidos was inspired by the work of Stephan and shared structural attributes with sulfonamido groups, namely a N–X=O fragment (X = S or P) that would fit into our tripodal ligand designs [59]. In contrast to sulfonamido groups, phosphinic amidos are more basic, and we reasoned that they would have similar basicity to the deprotonated urea groups in [H₃buea]³⁻. In addition, the P=O units in [poat]³⁻ would also promote the binding of a second metal complex and are stronger donors than S=O units, owing to their larger dipole moment. Therefore, the anionic, tetra-nitrogen binding pocket in [poat]³⁻ should provide strong σ-donation to access metal centers in the 4+ oxidation states and the P=O units can bind and stabilize auxiliary metal centers in their 3+ oxidation states.

2.1. Modulation of Fe^IV–oxido species using Lewis Acids and H-bonding

Our initial studies used [poat]³⁻ to prepare FeO(H) complexes which led to the formation of the high spin (S = 2) Fe^IV=O complex, [Fe^IVpoat(O)]⁻ (Fig. 6C) [54]. We were unable to crystallize any salt of this complex because of its thermal instability (it is only stable at temperatures below −60 °C). Nevertheless, working at these lower temperatures we were able to define the electronic structure using absorption, electron paramagnetic resonance (EPR) and Mössbauer spectroscopies, vibrational properties using nuclear resonance vibrational spectroscopy, and information about its molecular structure from X-ray absorption spectroscopy and Badger’s Rule [60]. We further discovered that [Fe^IVpoat(O)]⁻ can bind group II Lewis acids (LA^II = Mg, Ca, Sr, and Ba) near the Fe^IV–oxido unit to form [Fe^IVpoat(O)--LA^II]⁺ complexes (Scheme 2) [54]. The resulting electrostatic interactions had large effects on the electronic structure within the Fe^IV=O unit, enabling investigation on how such modulation influences structural and electronic properties within metal-oxido species. Specifically, our findings demonstrated that the binding of the Lewis acidic metal ions modulated the π-bonding within the Fe=O unit in which the acidity of the LA correlated with a weakening of the interactions between the Fe and oxido ligand.

Scheme 2.

General preparative route to heterobimetallic [Fe^IVpoat(O)--LA^II]⁺ complexes (LA^II = Mg, Ca, Sr, Ba).

A similar bonding effect was found with H-bonds which were established experimentally by comparing the properties of [Fe^IVpoat(O)]⁻ to those of [Fe^IVH₃buea(O)]⁻, a structurally similar high-spin Fe^IV–oxido complex that contains three intramolecular H-bonds between the oxido ligand and the tripodal ligand (see section 1.5). The Fe–O bond length in [Fe^IVpoat(O)]⁻ was shorter and its associated FeO vibration larger than those found in [Fe^IVH₃buea(O)]⁻; both results are consistent with the intramolecular H-bonds diminishing π-bonding [61]. Historically, multiple bonds have always been invoked to stabilize metal ion/oxido ion interactions; our results illustrate that in certain cases, H-bonds can supplant π-bonds in forming terminal metal-oxido complexes.

2.2. Formation of bimetallic complexes via a stepwise, site-directed approach

To further explore the multifunctionality of [poat]³⁻, we prepared a series of M^II–(μ-OH)–Fe^III complexes, where the M^II ion was varied across the period from Mn to Zn [62–65]. The monomeric starting synthon is a Fe^III–OH complex supported by [poat]³⁻, [Fe^IIIpoat(OH)]⁻, that can bind another transition metal complex through O atoms from the hydroxido ligand and two O atoms from the phosphinic amido groups of the [poat]³⁻ ligand. We also discovered that addition of pre-formed 1,4,7-trimethyl-1,4,7-triazacyclononane metal complexes ([M^II(TMTACN)]²⁺) to solution of in situ generated [Fe^IVpoat(O)]⁻ produced a bimetallic complex with a bridging oxido ligand (Scheme 3). This approach enabled formation of site-differentiated bimetallic complexes containing bridging hydroxido/oxido units where the identity of metal ions in each site can be systematically varied. The resulting species allows us to explore the individual effects of each metal ion on the electronic structures, redox properties, magnetic properties and reactivity of bimetallic cores.

Scheme 3.

General preparative route to [M^II–(μ-OH)–Fe^IIIpoat]⁻ cores (M^II = Mn, Fe, Co, Ni, Cu, Zn)

Within the context of ligand design, our approach melded concepts used by both Thomas and Wieghardt that were outlined in the first sections. The starting complex [Fe^IIIpoat(OH)]⁻ has an auxiliary metal binding site composed of three O atoms that is similar to the three phosphine units in the systems prepared by Thomas. The stepwise assembly to form M–(μ-O(H))–Fe cores, in which each metal has a different primary coordination sphere emulates what Wieghardt employed with different TACN ligands. In the remaining sections we will highlight some of our findings with FeFe, FeMn, and CoFe complexes to illustrate the validity of our initial rationalizations for the development high valent bimetallic species.

2.3. Reactivity of Fe^II–(μ-O)–Fe^IVpoat

Regardless of metal ion identity, a unifying structural motif among bimetallic dioxygen binding and activating enzymes is the presence of a μ-O or μ-OH core, where the single atom bridge preserves the integrity of the binuclear core throughout catalytic turnover and enables efficient magnetic exchange coupling between the two metal centers [66–73]. These traits are exemplified by metalloenzymes ribonucleotide reductases (RNRs) which utilize cofactors with diiron centers for the conversion of nucleotides to deoxynucleotides, a reaction that is essential for DNA biosynthesis in all domains of life [74]. For class Ia RNR, turnover proceeds through a series of oxidation steps starting from a Fe^IIFe^II core which ultimately generates an Fe^IV–(μ-O)–Fe^III species (intermediate X) that homolytically cleaves the O–H bond of a nearby tyrosine residue to form a tyrosyl radical to initiate the catalytic cycle (Scheme 4) [75–80]. Understanding the structural, electronic and magnetic properties of the FeFe core in RNR is critical for elucidating how nature stabilizes high-valent intermediates, orchestrates proton coupled electron transfer (PCET) processes, and couples radical chemistry to controlled substrate turnover. However, direct experimental interrogation of these fleeting intermediates spanning discreet oxidation steps remains challenging, motivating the development of synthetic diiron complexes that capture key features of the biological active sites, especially the generation of Fe^III–(μ-O)–Fe^IV cores like in intermediate X [81–84].

Scheme 4.

Proposed catalytic mechanism of RNR Class 1a

To this end, our group has recently reported a series of asymmetric [Fe–(μ-O(H))–Fe] complexes using [poat]³⁻, providing a platform to probe oxidation state changes, magnetic coupling, and proton dynamics relevant to the RNR mechanism [65]. In [Fe^II–(μ-OH)–Fe^IIIpoat]⁺ two of the phosphinic amido groups bind the second metal ion, while the remaining phosphinic amido group serves as an H-bond acceptor to stabilize the bridging hydroxido ligand (Scheme 5). One-electron oxidation of [Fe^II–(μ-OH)–Fe^IIIpoat]⁺ with [FeCp₂]OTf at −80 °C generates [Fe^III–(μ-OH)–Fe^IIIpoat]²⁺ (Table 2) Mössbauer and variable field temperature measurements corroborate the assignment of antiferromagnetically coupled S = 5/2 Fe^III centers, with a bridging hydroxido unit. Presumably the increased acidity of the hydroxido proton contributes to the thermal instability of [Fe^III–(μ-OH)–Fe^IIIpoat]²⁺ because at room temperature the oxido bridged complex [Fe^III–(μ-O)–Fe^IIIHpoat]²⁺ is the only species observed. The increased acidity leads to intramolecular proton migration to generate the oxido ligand and [Hpoat]²⁻, the mono-protonated form of [poat]³⁻ (Scheme 5). Protonation occurs at the phosphinic amido group that once served as the H-bond acceptor and this tautomer form still binds to the Fe center but now through the P=O unit. Mössbauer analysis suggests that this ligand protonation distorts the geometry around the fivecoordinate Fe^III center in [Hpoat]²⁻ while the auxiliary Fe^III center retains parameters similar to other six-coordinate Fe complexes. The antiferromagnetic exchange coupling between the high spin S = 5/2 Fe^III centers supports the di-Fe^III containing bridging oxido ligand (Table 2). This intramolecular proton transfer demonstrates the dynamic coordination sphere provided by the phosphinic amido/amide arms and its usefulness in controlling proton transfer.

Scheme 5.

Diiron complexes spanning four oxidation levels and various protonation states.

Table 2.

Summary of electrochemical, structural, and magnetic parameters for the series of bimetallic complexes of the type [M’–(μ-O(H))–Mpoat]⁺ [64,65,88].

[M’–(μ-O(H))–Mpoat]+	E l/2 a	M–Ob	M’–Ob	S M	S M′	S Total	J c
[FeII–(μ-OH)–FeIIIpoat]+	−0.42d	1.894	2.015	5/2	2	1/2	30
[FeII–(μ-OH)–MnIIIpoat]+	−0.38d	1.845	2.022	2	2	0	--
[CoII–(μ-OH)–FeIIIpoat]+	0.68e	1.890	1.993	5/2	3/2	1	17
[FeIII–(μ-O)–FeIIIpoat]+	0.55f	1.799	1.784	5/2	5/2	0	> 100
[FeIII–(μ-O)–MnIIIpoat]+	0.20f	1.767	1.802	3/2	5/2	1	> 100
[CoIII–(μ-O)–FeIIIpoat]+	--	1.784	1.788	5/2	2	1/2	> 100

^aV (vs Fe^III/^IICp₂);

^bÅ;

^ccm⁻¹;

^dM’^IIIM^III/M’^IIM^III;

^eE_a;

^fM’^IIIM^IV/M’^IIIM^III

While stable, [Fe^III–(μ-O)–Fe^IIIHpoat]²⁺ can be treated with base to produce [Fe^III–(μ-O)–Fe^IIIpoat]⁺, a pure salt that enables detailed characterization of the electronic and molecular structures of this species (Scheme 5). Again, Mössbauer spectroscopy revealed two distinct quadrupole doublets that are consistent with two high spin S = 5/2 Fe^III ions, with antiferromagnetic exchange coupling typical of an oxido bridge. This new species was found to be EPR silent, supporting an overall S = 0 ground state. Isolation of [Fe^III–(μ-O)–Fe^IIIpoat]⁺ further allowed for its use as a key synthon to access the higher valent Fe^IIIFe^IV species. Our electrochemical investigation found a reversible redox event at +0.55 V versus Fe^IIICp₂/Fe^IICp₂ that we assigned to the Fe^IIIFe^IV/Fe^IIIFe^III couple (Table 2). The detection of the desired Fe^IIIFe^IV analogue was realized by addition of a one-electron oxidant to [Fe^III–(μ-O)–Fe^IIIpoat]⁺ at −90 °C (Scheme 5). EPR and Mössbauer spectroscopies indicate the formation of [Fe^III–(μ-O)–Fe^IVpoat]²⁺, an S = ½ antiferromagnetically coupled system containing high spin Fe^III and Fe^IV centers, consistent with the Fe^III–(μ-O)–Fe^IV cores found in biological systems, including intermediate X in Class 1C RNR.

After detection and characterization of the high valent [Fe^III–(μ-O)–Fe^IVpoat]²⁺ species, its reactivity toward phenolic substrates was examined as functional probes of biological PCET processes to form phenoxy radicals. We found that [Fe^III–(μ-O)–Fe^IVpoat]²⁺, analogous to intermediate X in Class 1C RNR, can homolytically cleave O–H bonds of phenols through a PCET process. For example, addition of 2,6-di-tert-butyl-phenol (DTBP, BDFE_O–H(C₆H₆) = 77.0 kcal mol⁻¹) at −90 °C was monitored by EPR and UV-vis spectroscopies. The disappearance of the characteristic features of [Fe^III–(μ-O)–Fe^IVpoat]²⁺ coincided with the emergence of new features consistent with the formation of [Fe^III–(μ-OH)–Fe^IIIpoat]²⁺, indicating the reduction of the Fe^IV center accompanied by the protonation of the bridging oxido ligand. Upon warming to room temperature, this [Fe^III–(μ-OH)–Fe^IIIpoat]²⁺ species afforded the expected [Fe^III–(μ-O)–Fe^IIIHpoat]²⁺ complex whereby intramolecular proton transfer had occurred as described above.

Analysis of the organic reaction products by gas chromatography mass spectrometry (GCMS) analysis identified 3,3′,5,5′-tetra-tert-butyl[1,1′-biphenyl]4,4’-diol (bisphenol) and 3,3′,5,5′tetra-tert-butyldiphenoquinone (diphenolquinone), with a combined yield greater than 90 %. These products are diagnostic of a phenoxyl radical intermediate, supporting a mechanism in which [Fe^III–(μ-O)–Fe^IVpoat]²⁺ oxidizes DTBP via a PCET process to generate the reduced [Fe^III–(μ-OH)–Fe^IIIpoat]²⁺ species and the corresponding phenoxyl radical. Consistent with prior reports, the phenoxyl radical undergoes bimolecular homocoupling at the para-position to generate bisphenol, which can subsequently react with residual Fe^IIIFe^IV to form diphenolquinone. Taken together, these observations support a mechanistic model in which PCET occurs through proton transfer to the μ-O bridge concomitant with electron transfer to the Fe^IV center, resulting in net H-atom abstraction from the phenolic substrate. This PCET event is accompanied by structural reorganization within the ligand framework, where one phosphinic amido group of the [poat]³⁻ ligand serves as a secondary sphere proton relay, facilitating delivery of the proton to the redox active core. This synthetic system captures several essential features proposed for intermediate X in RNR, including a high valent Fe^III–(μ-OH)–Fe^IV core, metal-metal magnetic coupling, and the ability to mediate PCET through a proton transfer network. The involvement of ligand-based proton relays in stabilizing and directing PCET mirrors the role of second coordination sphere interactions in metalloproteins. More broadly, these findings demonstrate how synthetic bimetallic complexes can be used to dissect the interplay between electronic structure, proton movement, and reactivity in bioinspired high valent metal centers, offering insights that are otherwise difficult to access directly in enzymatic systems.

2.4. The heterobimetallic complex [Fe^III–(μ-O)–Mn^IVpoat]²⁺

In contrast to the diiron cofactor in RNR Class 1C, some other dioxygen binding/activating metalloproteins employ heterobimetallic MnFe active sites to carry out distinct chemical transformations. An illustrative example is the R2-like ligand binding oxidase (R2lox), which features a ligand set similar to RNR but catalyzes the oxidative crosslinking of valine and tyrosine residues rather than nucleotide reduction [85–87]. The divergence in function of R2lox from RNR Class 1C, despite near-identical primary coordination spheres underscores how subtle changes in metal composition and protein environment can dictate reactivity. In addition, the use of a heterobimetallic MnFe cofactor raises fundamental questions regarding 1) the structural determinants that favor Mn ion incorporation over Fe ion despite being contrary to the Irving-Williams series, 2) the functional advantages of a homobimetallic core relative to heterobimetallic alternatives, and 3) whether Mn enables access to reaction pathways that are inaccessible or suppressed in diiron systems.

Motivated by these considerations, coordination chemists have produced an array of complexes: the first point has recently been addressed by works of Tilley and Zhang [19,20]. Both have developed novel binucleating ligands to probe why Mn^II ions may bind preferentially over Fe^II ions. Our site-directed approach allows us to work around the challenges associated with the favorability of incorporation of different metal ions, so we have explored points 2 and 3 by constructing FeMn heterobimetallic complexes with [poat]³⁻ [88]. Analogous to the [Fe^II–(μ-OH)–Fe^IIIpoat]⁺ system, structural and spectroscopic evidence support the formation of [Fe^II–(μ-OH)–Mn^IIIpoat]⁺, and one-electron oxidation yields [Fe^III–(μ-O)–Mn^IIIHpoat]²⁺ (Table 2). The fully deprotonated complex [Fe^III–(μ-O)–Mn^IIIpoat]²⁺ can be prepared by treating [Fe^IIpoat]⁻ with O-atom transfer reagent isopropyl 2-iodoxybenzoate (IBX-iPr) to produce [Fe^IVpoat(O)]⁻ and subsequent addition of [Mn^IITMTACN]²⁺. Protonation of this oxido bridged species with 2,6-lutidinium triflate at −60 °C yields the hydroxide bridged of [Fe^III–(μ-OH)–Mn^IIIpoat]²⁺. Across the Fe^IIIMn^III series, EPR, UV-vis and Mössbauer spectroscopies support an S = 5/2 Fe^III site antiferromagnetically coupled to an S = 2 Mn^III center (Table 2). This coupling mirrors trends observed in the di-Fe analogues while introducing distinct electronic asymmetry arising from the Mn site.

Access to higher oxidation states was achieved by one-electron oxidation of of [Fe^III–(μ-OH)–Mn^IIIpoat]²⁺ with acetylferrocenium triflate, yielding of [Fe^III–(μ-O)–Mn^IVpoat]²⁺. Electronic absorption, EPR, and Mössbauer spectroscopies and variable-field measurements support the formation of [Fe^III–(μ-O)–Mn^IVpoat]²⁺ containing and S = 3/2 Mn^IV center antiferromagnetically coupled to an S = 5/2 Fe^III site (Table 2). While there are several examples of FeMn complexes with oxidation states below 4+, access to higher valent species has been lacking, underscoring the power of our modular synthetic approach.

Reactivity studies further reveal striking differences between the FeFe and FeMn systems. In contrast to the diiron analogue, [Fe^III–(μ-O)–Mn^IVpoat]²⁺ demonstrated no reactivity with DTBP under identical conditions. However [Fe^III–(μ-O)–Mn^IVpoat]²⁺ did react with less sterically hindered para-substituted phenols, suggesting that reactivity in the FeMn system is governed by a combination of substrate accessibility and redox thermodynamics. This trend points to a correlation between the more negative Mn^IV/III redox potential compared to Fe^IV/III and a reduced driving force toward homolytic cleavage of the phenolic O–H bond, consistent with a more controlled or selective PCET pathway relative to the diiron system.

Collectively, these results demonstrate that substitution of Mn for Fe within an otherwise conserved bimetallic framework significantly alters the physical properties of the complex, including spin states and substrate scope. In a biological context, such tuning may allow MnFe enzymes like R2lox to avoid possible undesirable Fe-based oxidative pathways, while enabling selective radical-mediated transformations such as amino acid cross-linking. These findings highlight how heterobimetallic active sites exploit metal identity to balance reactivity and control.

2.5. Electrochemical properties of [Fe–(μ-O(H))–Mpoat] where M=Mn or Fe

The distinct reactivity profiles of the FeFe and FeMn systems cannot be attributed solely to differences in structure or spin state, as both feature antiferromagnetically coupled high-spin metal centers bridged by μ-O(H) ligands. A useful framework for comparing the FeFe and FeMn poat systems is to begin from the Fe^IIIFe^III and Fe^IIIMn^III cores, which represent a shared structural and electronic reference point prior to accessing the lower or higher valent manifolds. Electrochemical measurements show that the Fe^IIIFe^III/Fe^IIFe^III and Fe^IIIMn^III/Fe^IIMn^III couples are essentially superimposable (both ~ −0.4 V), indicating that the Fe^III/II event is largely insulated from the identity of the second metal in this scaffold (Table 2) [65,88]. Divergence emerges when reaching the fully reduced states. The Fe^IIFe^III/Fe^IIFe^II couple occurs at −1.62 V, whereas the Fe^IIMn^III/Fe^IIMn^II is substantially less negative at −1.2 V and thus more easily reduced. While this difference does not directly alter reactivity, it may influence how electron density is partitioned within the dinuclear core which impacts the acidity of the μ-OH bridge and subsequent PCET steps. At the oxidizing extreme, the same pattern appears but in the opposite direction. For the μ-oxo bridged species, the Fe^IIIFe^IV/Fe^IIIFe^III couple is at +0.55 V whereas the corresponding Fe^IIIMn^III/Fe^IIIMn^IV couple is +0.20 V, indicating that Mn^IV is thermodynamically stabilized relative to Fe^IV (Scheme 6 and Table 2). However, the same stabilization that makes Mn^IV more accessible also limits the oxidizing power of the high-valent FeMn core relative to the FeFe.

Scheme 6.

Comparison of thermodynamic accessibility and reactivity of [Fe^III–(μ-O)–Fe^IIIpoat]²⁺ (A) versus [Fe^III–(μ-O)–Mn^IIIpoat]²⁺ (B).

This work provides a possible rationale for the evolutionary choices of metal ions: diiron cofactors may enable access to highly oxidizing Fe^IV intermediates necessary for long-range radical initiation, while MnFe cofactors stabilize high-valent states yet suppress unproductive or deleterious oxidation, favoring controlled, localized radical transformations. Furthermore, these findings illustrate how heterobimetallic architectures decouple structural requirements for dioxygen activation from thermodynamic control of downstream reactivity, offering a general strategy for tuning reactivity through deliberate manipulation of metal ion identity and redox potentials. The comparative analysis of the FeFe and FeMn systems established metal identity as a factor in shaping redox accessibility, PCET thermodynamics and overall reactivity within an otherwise conserved bimetallic framework.

2.6. Spin-state modulation enabled by Co substitution

While substitution of Fe with Mn within the [poat]³⁻ framework compresses the redox window and attenuates high valent oxidizing power, both FeFe and FeMn systems retain high spin metal centers across the accessible oxidation states. In these biologically inspired cores, changes in oxidation state modulate redox thermodynamic properties and reactivity without fundamentally altering the high spin nature of the individual metal ions. Moreover, while the magnetic exchange coupling increased in going from hydroxido to oxido bridges, the overall spin state of the complexes did not change. These observations prompted us to broaden our inquiry into what extent can the auxiliary metal ion be used to tune electronic communication and magnetic exchange, rather than just redox energetics, while preserving the integrity of the Fe^III(O(H))poat core. To address this question, we expanded our study to a systematic series of heterobimetallic complexes [M^II–(μ-OH)–Fe^IIIpoat]⁺ (M^II= Co, Ni, Cu, Zn) [62,64]. Within this series, the CoFe system emerges as a distinct system exhibiting redox- and proton-coupled changes in spin state at the Co center [64]. Electrochemical studies of [Co^II–(μ-OH)–Fe^IIIpoat]⁺ reveal an oxidative wave at E_pa = +0.68 V and a reductive feature at E_pc = −0.36 V (Table 2). Despite this apparent electrochemical irreversibility, the oxidized compound [Co^III–(μ-OH)–Fe^IIIpoat]²⁺ can be cleanly generated via one-electron chemical oxidation using AgOTf. The molecular structures of the complexes obtained from single crystal X-ray diffraction (XRD) methods show that the gross molecular structures of the Co^II and Co^III species are remarkably similar, indicating that the large potential separation (ΔE = 1.04 V) arises not from major structural rearrangement but from electronic reorganization coupled to changes in magnetism (Table 2).

Magnetic characterization highlights a fundamental departure from FeFe and FeMn chemistry. In [Co^II–(μ-OH)–Fe^IIIpoat]⁺, the high-spin Co^II center (S = 3/2) is antiferromagnetically coupled to Fe^III (S = 5/2), yielding an overall spin ground state of the complex, S_T = 1. Upon oxidation, the Co^III center becomes low spin and diamagnetic (S = 0), eliminating magnetic exchange with Fe^III and resulting in an overall S_T = 5/2 system dominated by the Fe^III site (Table 2). In contrast to FeFe and FeMn systems where oxidation preserves high-spin configurations, this redox event in the Co^III/IIFe^III complex directly switches the Co center from high spin to low spin, as expected for a six coordinate Co^III ion which favors a t_2g⁶e_g⁰ electron configuration.

The complex [Co^III–(μ-OH)–Fe^IIIpoat]²⁺ is resistant to the intramolecular proton migration at room temperature observed in the corresponding FeFe and FeMn analogues, prompting investigation into the deprotonation of the μ-hydroxido bridge. Treatment with a strong organic base (1,5,7-triazabicyclo[4.4.0]dec-5-ene, TBD) yields the oxido bridged species [Co^III–(μ-O)–Fe^IIIpoat]⁺, as confirmed by electronic absorption spectroscopy and structural determination from XRD methods (Fig. 7A,B). Deprotonation induces substantial elongation of Co–N bonds (~ 0.2 Å) and significant changes in Co–O distances, indicating an alteration in the Co electronic structure (Fig.7C,D). Magnetic and spectroscopic analyses reveal that deprotonation of the μ-OH bridge triggers an unprecedented spin-state inversion at the Co^III center. The low spin Co^III center in the hydroxido-bridged complex has switched to a rare high spin Co^III (S = 2) configuration in the oxido-bridged analogue. This paramagnetic, six-coordinate Co^III center is now strongly antiferomagnetically coupled to Fe^III (S = 5/2), producing an overall S_T = ½ ground state. This change is strikingly illustrated in the EPR changes associate with the deprotonation step (Fig. 7E,F): [Co^III–(μ-OH)–Fe^IIIpoat]²⁺ has a normal EPR spectrum for an S_T = 5/2 spectrum associate with the high spin Fe^III center, whereas the [Co^III–(μ-O)–Fe^IIIpoat]⁺ complex has a spectrum for the coupled S_T = ½ system with appreciable ⁵⁹Co hyperfine. The Fe^III center remains high spin in both complexes as determined by Mössbauer studies. Computational studies attribute this behavior to enhanced π donation from the μ-oxido ligand, which increases orbital overlap with d π* orbital of the Co^III center and increases their energy to such an extent that a high spin configuration is now stable. This electronic picture is consistent with crystallographic observations, where stronger π bonding is accompanied by shorten of the Co–O bond. However, the Co–N and remaining Co–O bonds are elongated because the σ* antibonding orbitals in [Co^III–(μ-O)–Fe^IIIpoat]⁺ are now populated.

Figure 7.

Deprotonation of hydroxo-bridged species [Co^III–(μ-OH)–Fe^IIIpoat]²⁺ (A) to produce oxo-bridged species [Co^III–(μ-O)–FeIIIpoat]+ (B) using TBD. Deprotonation leads to substantial elongation of Co–N bonds (C, D). [Co^III–(μ-OH)–Fe^IIIpoat]²⁺ has a normal EPR spectrum for an S_T = 5/2 system associated with a high spin Fe^III center (E), while [Co^III–(μ-O)–Fe^IIIpoat]⁺ has an EPR spectrum for the coupled S_T = ½ with Co hyperfine coupling (F).

These results establish a clear distinction between our FeFe and FeMn cores, which preserve high spin configurations across redox states, and the CoFe system, where redox events and proton transfer actively reprogram the spin state of the auxiliary metal ion. This behavior demonstrates that the Co ion enables redox- and proton-gated spin switching within an otherwise rigid bimetallic framework. The ability to reversibly modulate magnetic ground states through chemical stimuli positions the CoFe platform as a prototype for molecular spin switches, extending the relevance of poat-supported bimetallic complexes beyond enzymatic modelling into the realm of functional molecular magnetism and materials design.

3. Vistas

This review highlights how rational ligand design enables precise construction of bimetallic complexes in which metal identity, metal-metal separation, and bridging ligands can be independently manipulated to control function. We have highlighted our work which shows how proton and electron transfer can be regulated within homo- and heterobimetallic complexes that possess distinct primary coordination spheres. The [poat]³⁻ framework enables access to high-valent, oxo-bridged bimetallic motifs that capture key structural and functional elements of metalloenzyme cofactors. These studies further reveal the central role of the μ-oxido/μ-hydroxido ligand in modulating spin states at both individual metal centers and the total spin of the bimetallic core. Beyond bioinspired ligand design, the modular stepwise assembly strategy offers a general platform for expanding across a broader range of metal ions to map where substitution perturbs redox accessibility and/or magnetic exchange, and to identify where high-valent states or unusual spin manifolds become selectively accessible.

These insights echo principles observed in binuclear active sites in biology, where the presence of two or more metal centers is central to function. With this said, even in biological systems, the detailed connections between the local environment and function remain incompletely understood. This knowledge gap is exemplified by di-Fe cofactors involved in dioxygen chemistry. For instance, hemerythrin is a respiratory protein that functions as a reversible dioxygen binding protein, whereas a similar di-Fe cofactor in methane monooxygenases binds and activates dioxygen to convert methane to methanol, one of the most thermodynamically challenging transformations [89,90]. Bridging ligands and controlled metal/metal distances are central to function and bridging oxo/hydroxo ligands often serve dual roles as structural connectors and stabilizers for reactive intermediates, facilitating efficient electron and proton transfer during catalysis. Larger metal clusters are also found within active sites, and it appears that these cofactors have similarly evolved to promote the efficient transfer of electrons and protons. The oxygen-evolving complex (OEC) in photosystem II contains a Mn₄Ca cluster that mediates the four-electron oxidation of water to dioxygen, with each Mn center undergoing sequential one-electron oxidation events [91]. The OEC is surrounded by a large network of H-bonded water molecules which undoubtedly play an important role in the conversion of water to dioxygen.

Synthetic complexes that replicate these motifs therefore offer opportunities not only for understanding enzymatic mechanisms, but also for developing functional materials capable of controlled electron transfer, molecular magnetism, or redox-driven switching behavior. Ligand design provides a handle for controlling proton transfer pathways in addition to electron transfer. As demonstrated in our FeFe and FeMn poat systems, incorporating H-bonding groups, proton relays, or secondary sphere functional groups allows synthetic systems to mimic enzymatic strategies form managing PCET. Such control is particularly relevant for oxo- and hydroxo- bridged bimetallic cores, where the protonation state of the bridging ligand can influence function.

Several opportunities emerge from these findings. Still lacking in these systems are the connections thermodynamic properties with quantitative kinetic and mechanistic metrics for PCET and O–H bond activation. Systematic studies correlating redox tuning with substrate scope, reaction rates, and selectivity, together with deeper insight into how bimetallic motifs gate proton delivery and electron transfer, will be essential for developing predictive models for reactivity. The recurring role of the bridging ligand also suggests the μ-O(H) protonation state as a design parameter that can be exploited for the design of functional materials. The CoFe systems illustrate how protonation state and oxidation state can act as triggers to control magnetism, motivating exploration of stimuli-responsive spin switches driven by acids/bases, redox agents, or electrochemical stimulation. Embedding these cores into extended architectures could translate molecular level switching into functional applications for molecular magnetism.

While this review has focused primarily on systems involving transition metal ions, the incorporation of redox-inactive or Lewis acidic metal ions represents an important frontier. Again, nature provides an example in the Mn cluster within the OEC, where the Ca²⁺ ion is essential for function, though its role is still debated in literature. Strong synthetic advances have been made toward mimicking the OEC cofactor and the work of Agapie demonstrates how the clever designs of ligands can aid in the assembly of the cluster [92–94]. However, many putative heterobimetallic species involving main group and transition metal ions remain insufficiently characterized, particularly with respect to structural and mechanistic details of catalytically relevant intermediates. Many examples where the reactivity of transition metal complexes is enhanced by a strongly Lewis acid main group metal ion, such as Sc³⁺, Al³⁺, and Mg²⁺ ions have been reported, but the potential for Lewis acidic metal ions to function as Brønsted acids via their aqua/hydroxido complexes is often underappreciated [95–97]. It is here that more work is needed to at least determine the formulations of key intermediates. Advancing this area in both inorganic and organic catalysis will require new ligand platforms capable of selectively and reproducibly binding metal ions with distinct coordination preferences, while maintaining sufficient rigidity to promote intramolecular cooperativity and suppress undesired intramolecular processes.

The tunable magnetic properties of bimetallic complexes also present opportunities for practical applications in sensing. Most molecular sensors currently rely on mononuclear complexes, particularly Fe^II systems that exhibit large spin-state changes in response to an analyte [98–100]. In contrast, complexes containing an isoelectronic Co^III have never been utilized because it is thought that there were no existing ligands that could support both high and low spin states. Our studies on the [Co^III–(μ-O(H))–Fe^IIIpoat]^2+/+ system establish that Co^III centers can, in fact, be harnessed through appropriate ligand design to achieve switchable magnetic behavior [64]. The protonation state changes at a bridging oxido/hydroxide ligand can trigger spin state changes at the Co^III center that lead to large differences in the paramagnetism of the bimetallic complex [64]. These findings suggest that bimetallic systems represent a promising and largely underexplored platform for achieving responsive magnetic materials.

Highlights:

• Rational ligand design enables site-directed assembly of bimetallic complexes.

• Modular bimetallic platforms reveal how metal identity, redox thermodynamics, and protonation states cooperatively govern reactivity and spin-state behavior

• Synthetic FeFe and FeMn complexes capture key features of enzymatic cofactors

• Redox and proton-gated spin switching in CoFe complexes establishes bimetallic cores with tunable molecular magnetic units