Synthesis and Applications of Perfunctionalized Boron Clusters

Abstract

This Viewpoint Article describes major advances pertaining to perfunctionalized boron clusters in synthesis and their respective applications. The first portion of this work highlights key synthetic methods allowing one to access a wide range of polyhedral boranes (B₄ and B₆ – B₁₂ cluster cores) that contain exhaustively functionalized vertices. The second portion of this Viewpoint showcases the historical developments in using these molecules for applications ranging from materials science to medicine. Lastly, we suggest potential new directions for these clusters as they apply to both synthetic methods and applications.

Graphical abstract

The present Viewpoint Article summarizes the syntheses and applications of perfunctionalized boron clusters with particular highlights on foundational works as well as the most up-to-date examples. We also provide a brief outlook on the field as it stands in the context of potential further development and application of boron cluster chemistry to solve pressing problems in chemistry, materials science and medicine.

Introduction

Carbon-based chemistry has been at the forefront of chemical research and has occupied the thoughts and efforts of its leading exponents for over two centuries. Despite its proximity to carbon in the periodic table, boron-based chemistry (though rich) has lagged far behind its first-row neighbor and in most cases has been used in service of carbon-based chemistry,¹ rather than as a building block in its own right.

For boron chemistry, the history of borane (or boron hydride) chemistry began with the pioneering work of Alfred Stock, who with his co-workers isolated a series of simple boranes, including the classic diborane(6) B₂H₆.² However, it was not until the theoretical prediction of the icosahedral borane [B₁₂H₁₂] by Lipscomb and co-workers in 1954 that the unusual nature of the bonding in boranes become apparent. Here, the concept of boron clusters was introduced,³ with subsequent studies by Longuet-Higgins and Roberts suggesting that the stable form of this icosahedron would have an overall 2- charge.⁴ These theoretical studies were then validated by Hawthorne and Pitochelli in a series of reports highlighting the first synthesis of dodecaborate [B₁₂H₁₂]²⁻ as well as decaborate [B₁₀H₁₀]²⁻ species.^5,6

Contemporary books and reviews have broadly showcased recent advances in boron cluster chemistry and their applications,^7–19 but in this Viewpoint we hope to reveal to the wider inorganic chemistry community the immense potential of boron clusters by focusing on a particular class of molecules, namely perfunctionalized clusters whose cores are composed entirely of boron atoms. We will discuss their synthesis and unique attributes and then present examples of their use in some unusual and unexpected arenas, ranging from photo-activation of olefins to hybrid nanomolecules. Finally, we will discuss potential future directions for these unique molecules and hope to provide inspiration for the wider community to engage in boron cluster chemistry.

Definition of a Cluster in this Viewpoint

In this Viewpoint, we will be discussing the synthesis and applications of perfunctionalized all-boron clusters. Here the clusters are in their closo polyhedral forms as defined by Wade,²⁰ where n number of boron atoms contain (n+1) skeletal bonding electron pairs, typically resulting in anions of the type closo-[B_xH_x]²⁻ (x = 6 – 12 and above). These molecules exhibit unique bonding situations characterized by multicenter, two electron bonding and three-dimensional aromaticity and electron delocalization. These characteristics are believed to contribute to the high stability of these species and have been reviewed in the literature.²¹ This bonding situation is qualitatively visualized in 1) one set of atomic orbitals whose radial extension toward the center of the cluster results in the in-phase overlap of n boron-based orbitals of a cluster of the general formula B_xH_x^y−, forming a molecular orbital in the center of the cluster shared by all boron atoms and 2) one set of atomic orbitals whose in-phase contributions result in a molecular orbital delocalized over the surface of the pseudo-spherical cage (Fig. 1A–B). For thoroughness, some perfunctionalized boron cluster classes will be addressed that deviate from Wade’s rules, such as the 4-membered polyhedral boron clusters and charge-neutral boron subhalides that have been reported. Furthermore, in addition to the expected closo-[B_xH_x]²⁻ boron cluster forms (x = 6 – 12), we will show that perfunctionalization of such molecules can also lead to hypocloso-[B_xH_x]¹⁻ and hypercloso-[B_xH_x]⁰ derivatives. Insofar as the parentage of these monoanionic and neutral species lie in boron clusters that abide by Wade’s rules, they will be included in this Viewpoint.

Figure 1.

Calculated structures (based on X-ray crystal structures) and representative Kohn-Sham molecular orbitals of A_1g symmetry demonstrating the delocalization of electron density in A. closo-[B₆H₆]²⁻ and B. closo-[B₁₂H₁₂]²⁻ as described by King.²¹ This bonding arrangement gives rise to high kinetic stability relative to tri-coordinate and non-polyhedral boranes.

Synthesis of Perfunctionalized Boron Clusters

B₄ and B₆ Clusters

The smallest boron-based clusters known are the deltahedral tetraboranes.^22,23 In contrast to higher-membered congeners (e.g., B₆ or B₁₂), these clusters do not display the expected 2n+2 bonding electron count that is commonly encountered in 3D-aromatic boron clusters.^21,24 The first known boron tetrahedrane, tetraboron tetrachloride,²⁵ was reported in 1952 as a thermal decomposition product of B₂Cl₄ (Fig. 2), obtained only in small yields; its solid-state X-ray structure was subsequently solved in 1953.²⁶ The early procedures for synthesizing this molecule involved either electrical or microwave discharges in the presence of BCl₃ and mercury,²³ with more recent procedures that produce workable yields involving the reduction of R–BF₂ fragments by Na/K alloy in alkane solvents.^27,28 Alternatively, B₄R_nCl_4-n (n = 1, 2, 4) analogues have been synthesized through treatment of the halogenated B₄Cl₄ cluster with alkyllithium reagents.^29,30 The reactivity of B₄Cl₄ has been described by Morrison and is summarized in Fig. 2.³⁰

Figure 2.

Summary of reactivity of tetrachloro-closo-tetraborate.

From haloaminoborane precursors, Siebert and co-workers synthesized tetraborane compounds of the type B₄(NR₂)₄ (R = Me, Et, ⁱPr, or NR₂ = 2,2,6,6-tetramethylpiperidinyl (TMP)).³¹ However only in the cases of R = Et or TMP were closo-tetraborane products obtained; the other derivatives were isolated as cage-opened species. Mention has also been made of a tetramesityltetraborane, though to our knowledge this molecule has not been published.³² Finally, Paetzold and co-workers have reported the synthesis of neopentyl and mixed alkyl tetraborate derivatives, however these neutral compounds were obtained in yields of 5% or less.³³

In comparison to larger cluster congeners, the functionalization and applications of 6-membered, closo-[B₆H₆]²⁻ clusters is largely underdeveloped. Most of chemistry conducted with these species was performed by Preetz and co-workers, who showed that in contrast to the typical Lewis acidity and high electrophilicity of tricoordinate boron, the hexaborate dianion is in fact nucleophilic at boron. It is generally accepted that this nucleophilicity is a result of the maxima of electron-density residing at each face of the hexaborate octahedron. A Microreview by Preetz and Peters describes the majority of the synthesis and substitution chemistry of closo-[B₆H₆]²⁻ clusters.³⁴ It is noteworthy, however, that the issue of hexaborate peralkylation was addressed briefly in this account, in which it was suggested that only up to three vertex substitutions by alkyl substituents were possible as a consequence of the steric crowding of alkyl substituents and difficulty in facial proton removal prior to further substitution (vide supra). We have recently discovered, however, that under appropriate conditions, perfunctionalization is in fact possible (vide infra).

The first mention of a perfunctionalized closo-[B₆H₆]²⁻ species was made by Klandberg and Muetterties in 1966.³⁵ However, not until almost 20 years later was a directed preparation of perhalogenated hexaborate dianions disclosed by Preetz and Fritze via treatment of Na₂[B₆H₆] with halogen in basic, aqueous media to produce closo-[B₆X₆]²⁻ (X = Cl, Br or I).³⁶ The resulting compounds were thoroughly studied by IR and Raman spectroscopy,³⁷ as well as ¹¹B NMR spectroscopy, which revealed an upfield shift of boron resonances from closo-[B₆Cl₆]²⁻ to closo-[B₆Br₆]²⁻ to closo-[B₆I₆]^2−.38 Subsequently, it was discovered that halogenation could be performed using succinimide-based reagents, however these conditions were employed largely for synthesizing only partially halogenated derivatives (Fig. 3).³⁹

Figure 3.

Reactivity of closo-[B₆H₆]²⁻, as described in the literature.

In cases of either diatomic halogen or NXS-based halogenation, species of the type closo-[B₆X_nH_6-n]²⁻ (n = 1 – 5) were generally observed, opening the possibility for heteroleptic perhalogenated closo-hexaborates. Interestingly, the synthesis of the disclosed bis-heteroleptic perhalogenated hexaborates, which form under analogous conditions to homoleptic perhalogenated derivatives, require a specific, sequential treatment with halogen, in which the more electronegative halogen must first be charged to the hexaborate cluster, followed by the less electronegative halogen.^39,40 In this way, halogenated hexaborate clusters of the types closo-[B₆Cl_nBr_6-n]²⁻, closo-[B₆Cl_nI_6-n]²⁻, and closo-[B₆Br_nI_6-n]²⁻ may be synthesized. To the best of our knowledge, the chemistry of these bis-heteroleptic clusters has not been further explored.

Prior to our involvement, the only non-halogen containing perfunctionalized six-membered closo-boron clusters were the neutral hypercloso-B₆(NEt₂)₆,^41,42 with Berndt and co-workers also disclosing the synthesis and structure of the dimethyl analogue, hypercloso-B₆(NMe₂)₆.⁴³ Given our interest in the development of chemistry and applications of perfunctionalized boron clusters,^44–46 we reexamined the reactivity of the closo-[B₆H₇]¹⁻ anion. Under microwave heating conditions, the hexaborate anion can be substituted with benzyl bromide electrophiles (Fig. 4A), affording a perfunctionalized cluster of the type closo-[B₆(CH₂Ar)₆H^fac]¹⁻ (Ar = C₆H₅, 4-Br-C₆H₄), containing six newly formed B–C bonds. Importantly, these cluster molecules can be isolated as moisture and air-stable solids. Crystallographic characterization of one derivative confirmed the formation of six B–C linkages. Interestingly, in contrast to the perhalogenated hexaborates reported by Preetz (vide supra), these alkylated clusters are marked by their relative oxidative instability, distinguishing them from the known redox-reversible substituted dodecaborate clusters (vide infra). Overall, it is clear that the original observations by Preetz suggesting that closo-[B₆H₆]²⁻ cluster features nucleophilic reactivity at boron sites can be extended towards building complex perfunctionalized cluster scaffolds.

Figure 4.

A. Synthesis of perfunctionalized B₆ clusters featuring direct B–C bonds. B. X-ray crystal structure of closo-[B₆(CH₂-4-Br-C₆H₄)₆H^fac]¹⁻.

B₇–B₁₁ Clusters

Moving beyond the small octahedral B₆ cluster, several additional three-dimensional B_nH_n²⁻ clusters (n = 7 – 11) have been studied,^6,47,48 though with the exception of B₁₀ decaborate clusters, there has been relatively little exploration of their functionalization. The perfunctionalization chemistry of closo-[B₁₀H₁₀]²⁻ was developed somewhat in parallel to that of the closo-[B₁₂H₁₂]²⁻ dodecaborate cluster in the early 1960s,^49–53 with several routes to diverse functionalized derivatives resulting in partial substitutions of the cage-bound hydrogens.⁵⁴ Despite this early progress, in the subsequent decades only halogenation resulted in successful substitution of all B–H vertices to achieve perfunctionalization of the B_nH_n²⁻ (n = 7 – 11) clusters to form species of the type B_nX_n²⁻ (X = Cl, Br and I).^{30,53,55–59}

In 1964, the first fully halogenated cluster in this series was the decaborate dianion, with [B₁₀Cl₁₀]²⁻, [B₁₀Br₁₀]²⁻, and [B₁₀I₁₀]²⁻ all synthesized by treating [B₁₀H₁₀]²⁻ with chlorine, bromine, and iodine, respectively.⁵³ Several other perhalogenated clusters were synthesized via the thermal decomposition of diboron tetrahalide B₂X₄ or boron trihalide BX₃, thus forming the perhalogenated clusters directly from these non-polyhedral precursors⁵⁵ rather than subjecting the boron hydride B_nH_n clusters to halogens directly. There is some mass spectral evidence suggesting that B₉(t-Bu)₉ was formed from the reaction of [B₈Cl₈]²⁻ with tert-butyl lithium,⁶⁰ but the supposed product was neither isolated nor fully characterized.

A more exhaustive history of the functionalization chemistry possible with these clusters that fall short of full persubstitution can be found in excellent reviews by Morrison³⁰ (for B_7–9,11) and Sivaev⁵⁴ (for B₁₀). The limited perfunctionalization observed over the past few decades can be largely attributed to lower stability of these clusters compared to closo-[B₁₂H₁₂]²⁻, which precludes harsher functionalization strategies. The stability of the boron subhalide clusters was predicted from their valence state index to be 9>11>8>10≈7,⁵⁶ though experimental thermal stability of the neutral clusters B_nX_n was found to be 9>10>8≈7,⁵⁵ showing poor agreement between the two methods but confirming the significantly lower stability of these clusters compared to [B₁₂H₁₂]²⁻. More recent investigations of these clusters have resulted in the X-ray crystallographic analysis of the structures as well as further functionalization, however the halogenated derivatives remain the only perfunctionalized clusters in the B_nH_n (n = 7–11) series to date.^57,61–64

B₁₂ Clusters

Due to its stability and commercial availability, the vast majority of reports on perfunctionalized boron clusters concern the deltahedral closo-dodecaborate dianion, closo-[B₁₂H₁₂]²⁻. For clarity, the following sections will be partitioned into those concerning homoperfunctionalized (Fig. 5) and heteroperfunctionalized (Fig. 6) B₁₂-based clusters.

Figure 5.

Homoperfunctionalization of closo-[B₁₂H₁₂]²⁻. Cations omitted for clarity. a. Alkylation. Peralkylation of dodecaborate can be achieved by refluxing [NⁿBu₄]₂[B₁₂H₁₂] with neat AlMe₃ and iodomethane over eleven days to form [NⁿBu₄]₂[B₁₂Me₁₂].⁸⁰ This remains the only reported peralkylation of dodecaborate, and the only homopersubstituted dodecaborate that contains B–C linkages. b) Halogenation^52,65–68 c) Hydroxylation^69–71,81 d) Carbamation⁸² e) Carbonation⁸² f) Esterification⁷² g) Etherification⁷¹.

Figure 6.

Established routes for the heteroperfunctionalization of closo-[B₁₂H₁₂]²⁻. Cations have been omitted for clarity.

Homoperfunctionalized B₁₂ Clusters

Halogenation

The first reported perfunctionalization of closo-[B₁₂H₁₂]²⁻ involved direct perhalogenation with elemental chlorine, bromine and iodine, resulting in molecules of the type closo-[B₁₂X₁₂]²⁻ (X = Cl, Br or I) that could be handled without decomposition under ambient conditions.⁵³ To avoid the use of chlorine gas, Ozerov and co-workers recently showed that perchlorination of closo-[B₁₂H₁₂]²⁻ may be achieved using SO₂Cl₂ as a chlorinating agent, leading to a more operationally simple synthesis.⁶⁵ Perfluorination of closo-[B₁₂H₁₂]²⁻ proved more difficult and was first achieved by Solntsev and co-workers in the early 1990’s through the action of supercritical HF at 550°C, producing Cs₂[B₁₂F₁₂] in 38 % yield.⁶⁶ Later, work by Strauss and co-workers described improved yields of closo-[B₁₂F₁₂]²⁻ salts up to 74 % and could be carried out on multigram scale using non-specialized glassware.^67,68

Hydroxylation

Hawthorne and co-workers reported the first successful perhydroxylation of dodecaborate by refluxing Cs₂[B₁₂H₁₂] in 30 % H₂O₂ for several days to form Cs₂[B₁₂(OH)₁₂].⁶⁹ Through cation exchange protocols, the B₁₂(OH)₁₂²⁻ dianion can be isolated as a variety of metal, acid, or organic salts M₂[B₁₂(OH)₁₂] (where M = Li⁺, Na⁺, K⁺, Rb⁺, Cs⁺, [H₃O]⁺, [NH₄]⁺, [NⁿBu₄]⁺, [(PPh₃)₂N]⁺).^70,71 Unlike their organic analogues, most perhydroxylated dodecaborate salts show low solubility in water, with only Cs₂[B₁₂(OH)₁₂] showing appreciable solubility and [NⁿBu₄]₂[B₁₂(OH)₁₂] showing excellent solubility at slightly elevated temperatures. In addition, the tetra-n-butyl ammonium salt is soluble in organic solvents, making it a valuable precursor for further derivatization.

Esterification

[NⁿBu₄]₂[B₁₂(OH)₁₂] can be transformed into esters of the type [NⁿBu₄]₂[B₁₂(OC(O)R)₁₂] (R = Me or Ph) using acetic anhydride or benzoyl chloride, as reported by Hawthorne and co-workers. Later reports from the same laboratory further expanded the scope of substituents that could be installed on the B₁₂(OH)₁₂²⁻ core through ester linkages.^72–74

Etherification

closo-[B₁₂(OH)₁₂]²⁻ reacts with a variety of alkyl or benzyl halides in the presence of base to form per-etherated clusters. The first such compound reported involved the reaction of [(PPh₃)₂N]₂[B₁₂(OH)₁₂] with benzyl chloride and ⁱPr₂NEt in refluxing acetonitrile for several days to form (after workup) Na[(PPh₃)₂N][B₁₂(OCH₂Ph)₁₂] in 48 % yield.⁷⁵ A variety of other peretherated clusters were also reported using similar conditions.⁷¹ Although these methods were effective, the length of time required for the reactions reduced their utility. To address this issue, our laboratory has reported a rapid microwave-assisted method to prepare peretherated clusters.⁴⁴ This method drastically reduces the reaction time from an average of multiple days to minutes to hours. Etherification of the perhydroxylated clusters has a profound effect on the redox properties of B₁₂-clusters and is discussed separately in the text (vide infra).

Heteroperfunctionalized B₁₂ clusters

B–X/B–N functionalization

While Muetterties and co-workers reported a variety of B₁₂ clusters containing B–X and B–N bonds as early as 1965, none of the molecules described were perfunctionalized.⁷⁶ Strauss and co-workers were successful in perfluorinating closo-[H₃NB₁₂H₁₁]¹⁻ to form closo-[H₃NB₁₂F₁₁]¹⁻ in situ, which could then be isolated via precipitation as the ammonium salt [NⁿBu₄]₂[H₂NB₁₂F₁₁].⁷⁷ An important feature of closo-[H₃NB₁₂F₁₁]¹⁻ is the reduction of the overall charge on the cluster from 2- to 1- as a result of charge compensation by the ammonium fragment. In principle, this allows the cluster to be treated as an analogue of the significantly more expensive polyhedral carba-closo-dodecaborate closo-[CB₁₁F₁₁]¹⁻, an exceptional weakly coordinating anion. Further functionalization of closo-[H₃NB₁₂F₁₁]¹⁻ at nitrogen affords dodecaborate anions of the type closo-[R₃NB₁₂F₁₁]¹⁻ (R = Me and n-C₁₂H₂₅). Jenne and co-workers were successful in the synthesis of the perchlorinated derivative closo-[H₃NB₁₂Cl₁₁]¹⁻ using SbCl₅ as a chlorinating reagent,⁷⁸ with further substitution then carried out at nitrogen to produce closo-[Me₃NB₁₂Cl₁₁]¹⁻. Duttwyler and co-workers later reported a simplified synthesis whereby SbCl₅ was replaced with SO₂Cl₂.⁷⁹

B–X/B–O functionalization

Muetterties and co-workers reported the first example of a mixed halogenated/hydroxylated B₁₂ cluster, closo-[B₁₂F₁₁OH]²⁻, as a byproduct of the attempted perfluorination of closo-[B₁₂H₁₂]²⁻ with elemental fluorine in water.⁵³ Jenne and Kirsch reported the synthesis of the heteroperfunctionalized cluster closo-[B₁₂X₁₁OH]²⁻ and subsequent alkoxylation to form a variety of compounds of the type closo-[B₁₂X₁₁OR]²⁻ (where X = Cl or Br; R = n-C₃H₇, n-C₈H₁₇ or n-C₁₂H₂₅).⁸³ As with the formation of closo-[H₃NB₁₂Cl₁₁]¹⁻, Duttwyler and co-workers described a simplified synthesis of closo-[B₁₂Cl₁₁OH]²⁻ using SO₂Cl₂ instead of chlorine gas and showed additional reactivity at the hydroxyl group to produce closo-[B₁₂Cl₁₁O(SO₂CF₃)]²⁻ and closo-[B₁₂Cl₁₁OTs]²⁻ respectively.⁸² A summary of these functionalization methods is shown in Figure 5.

B–N/B–O functionalization

Hawthorne and co-workers synthesized the mixed B–N/B–O functionalized cluster closo-[B₁₂(OH)₁₁(NH₃)]¹⁻ via amination of closo-[B₁₂H₁₂]²⁻ to form closo-[B₁₂H₁₁(NH₃)]¹⁻ followed by hydroxylation to form closo-[B₁₂(OH)₁₁(NO₂)]²⁻ and a final reduction step with Raney nickel to afford the desired closo-[B₁₂(OH)₁₁(NH₃)]¹⁻ species.⁸⁴ The authors report that nitro derivative closo-[B₁₂(OH)₁₁(NO₂)]²⁻ may be used directly for further functionalization of the hydroxyl group, with subsequent reduction to yield products of the type closo-[B₁₂(OR)₁₁(NH₃)]¹⁻ (see Figure 6).

Redox Chemistry of Perfunctionalized Clusters

Upon perfunctionalization, boron clusters can access electronic states that do not conform to Wade’s rules. This gives rise to molecules that potentially have access to three distinct oxidation states: closo-[B_xR_x]²⁻, hypocloso-[B_xR_x]¹⁻ and hypercloso-[B_xR_x]⁰. For hexaborate clusters, this redox chemistry is readily accessible for the homoleptic perhalogenated cluster derivatives closo-[B₆X₆]²⁻ (X = Cl, Br or I). Use of choice oxidants sees their transformation to their highly colored open shell radical monoanions, hypocloso-[B₆X₆]¹⁻ (X = Cl, Br or I).^85,86 In comparison to the oxidation potential of the parent closo-[B₆H₆]²⁻ (0.86V vs. Ag/AgCl/LiCl), closo-[B₆Cl₆]²⁻ and closo-[B₆Br₆]²⁻ are more easily oxidized, with oxidation potentials lying at 0.58V and 0.77V, respectively; in contrast, closo-[B₆I₆]²⁻ demonstrates approximately the same oxidation potential as closo-[B₆H₆]²⁻ at 0.88V. The perhalogenated derivatives B₉X₉ (X = Cl, Br, I) were found to have two highly reversible, well-separated one-electron steps between the dianionic and neutral states, with significantly lower oxidation potentials compared to the B₁₂X₁₂ species (−0.63 V to −0.74 V vs Fc/Fc⁺ for the 2- to 1- oxidation step for B₉X₉ clusters, compared to 1.68 V to 2.27 V vs Fc/Fc⁺ for the same redox event with B₁₂X₁₂).^87,88

As mentioned, the vast majority of reports on perfunctionalized boron clusters concern the 12-vertex polyhedral closo-[B₁₂H₁₂]²⁻. This particular class of molecule has a rich redox chemistry that has only recently been exploited. Without modification, boron hydride clusters such as closo-[B₁₂H₁₂]²⁻ have very high redox potentials, which can result in irreversible cluster degradation or the formation of B–B linked dimers when subjected to controlled oxidiation.⁸⁹ However, perfunctionalization of the closo-[B₁₂H₁₂]²⁻ scaffold to form halogenated or other derivatives engenders reversible redox behavior.¹¹ Early examples of this phenomenon include reports that the perhalogenated dodecaborates could undergo single electron oxidation, forming a stable hypocloso-[B₁₂X₁₂]¹⁻ radical.⁹⁰ Attempted isolation of one such species, [B₁₂Cl₁₂]¹⁻, instead resulted in the isolation of neutral B₁₂Cl₁₂, evidencing the structural stability but potentially high reactivity of these radicals.⁹¹ However, other dodecaborate radicals, such as salts of perhydroxylated closo-[B₁₂(OH)₁₂]^{1− 92} as well as permethylated [B₁₂Me₁₂]^{1− 93} have been isolated as stable anions. Their stability is largely due to the 3-dimensional delocalization of electron density throughout the cluster core, which mitigates the potent reactivity of a localized radical (Fig. 7). Furthermore, steric protection of the cluster core through B-bound substituents also kinetically stabilizes the hypocloso species.

Figure 7.

A. X-ray crystal structure of [(PPh₃)₂N][B₁₂Me₁₂] ([(PPh₃)₂N]⁺ counterion omitted for clarity). B. The singly occupied molecular orbital (SOMO) and the spin density localization (PBE/D3BJ:TZP).

More recent explorations of this reversible redox behavior have been conducted with perfunctionalized derivatives of closo-[B₁₂(OH)₁₂]²⁻. It was demonstrated that the ether-linked benzyl closo-[B₁₂(OCH₂Ph)₁₂]²⁻ cluster can undergo two sequential, quasi-reversible one-electron oxidation reactions, enabling access to three distinct redox states: closo-[B₁₂(OCH₂Ph)₁₂]²⁻, the stable radical hypocloso-[B₁₂(OCH₂Ph)₁₂]¹⁻, and the neutral hypercloso-B₁₂(OCH₂Ph)₁₂. In all three oxidation states, the dodecaborates were stable, isolable products that were fully structurally characterized.⁷⁵ The most useful feature of this redox behavior was the discovery that the individual redox potentials of the closo-[B₁₂(OR)₁₂]²⁻ clusters could be rationally tuned — in a manner reminiscent of metal-based inorganic complexes — as a function of the O-bound alkyl or benzyl substituent.^11,71,94 Recent work has demonstrated even higher oxidation potentials for some newer closo-[B₁₂(OR)₁₂]²⁻ derivatives, continuing the correlation between the predicted effects of the substituents (Fig. 8) while expanding the tunable window for the same redox event to a full 1 V (−0.61 V vs Fc/Fc⁺ for the 1-/0 redox couple oxidation when R = n-hexyl, rising to +0.68 V vs Fc/Fc⁺ for the same redox event when R = (3,5-bis)trifluoromethylbenzyl).^44,94

Figure 8.

A. Redox potentials of [B₁₂(OR)₁₂] clusters plotted vs Hammett constants (top/blue = 1- to 0, bottom/red = 2- to 1-). B. Single crystal X-ray structure of cluster [e]¹⁻. C. UV-vis spectra and photos of the air-stable radical cluster [e]¹⁻ (red line) and the dianionic cluster [e]²⁻ (blue dashed line). D. EPR spectrum of the radical cluster [e]¹⁻. These highlight the tunable nature of the perfunctionalized clusters.^10,44,95

Applications in Catalysis and Materials Chemistry

While boron clusters may have existed as chemical curiosities over the last century, they became a topic of intense interest for scientists during the 1950s, who, at the height of the space race, were keen to apply them towards the development of ever more powerful energetic materials for rocket fuels. Since interest quelled in the use of boron clusters in that arena, applications for boron clusters have been few and far between. Having outlined the state-of-the-art with regards to the synthesis of perfunctionalized boron clusters and extolled the virtues of their signature features, in this section of the Viewpoint we will highlight select examples of areas in which perfunctionalized clusters may be exploited.

Photoredox Behavior

Perfunctionalized carba-closo-dodecaborates of the type closo-M[CB₁₁X₁₁] have found exceptional utility in homogeneous catalysis due to their ability to act as inert, extremely weakly coordinating anions.^96,97 While dodecaborates composed entirely of boron, which bear a charge of 2-, may be reminiscent of these weakly coordinating carboranes, the ability to functionalize these molecules and the properties that manifest as a result present a powerful opportunity to move perfunctionalized boron clusters away from a spectator role and towards being an active participant in reaction chemistry. In particular, the unique ability to tune the redox window of dodecaborates through perfunctionalization via judicious choice of functional group (vide supra) provides straightforward access to air- and moisture-stable molecules that have well-defined one-electron redox chemistry across a range of potentials. Such redox chemistry could then be exploited to mediate a plethora of transformations across a range of substrates, irrespective of substrate reactivity.

During the course of our studies on improving the synthesis of ether linked perfunctionalized clusters,⁴⁴ we observed that solutions of hypercloso-B₁₂(OCH₂Ph)₁₂ left to stand in ambient light in the presence of with 4-methoxystyrene led to the formation of viscous mixtures; in the dark, no such mixture was obtained. By utilizing blue LEDs as a light source under controlled conditions, we demonstrated that the perfunctionalized cluster hypercloso-B₁₂(OCH₂Ph)₁₂ can behave as a powerful photooxidant in the visible light-assisted polymerization of olefins.⁴⁶ The proposed mechanism involves the generation of a potent photooxidant by visible light promotion of an electron from a low lying occupied orbital localized largely on the aryl rings and oxygens to the cluster-based LUMO (Fig. 9A). Unlike metal-initiated polymerizations of olefins which generally require an open coordination site for monomer approach, these clusters are devoid of well-defined coordination sites. It was therefore hypothesized that favorable, non-covalent interactions of the monomer and the aryl ring(s) of the benzyl substituents on the cluster periphery facilitate electron transfer from monomer to the dodecaborate core (Fig. 9B). Time-dependent Density Functional Theory (TD-DFT) studies carried out on both hypercloso-B₁₂(OCH₂Ph)₁₂ and hypercloso-B₁₂(OCH₂C₆F₅)₁₂ suggest that this charge-transfer excitation pathway is favorable and leads to an excited state species with a redox potential sufficiently anodic for the one-electron oxidation of styrene. Importantly, the energy of this excitation calculated in silico corresponds to the energy provided by a photon of blue (~450nm) light, which is consistent with the ~450nm absorption band observed for the benzyloxy-substituted hypercloso-B₁₂-based species (Fig. 9C). In order to render more electronically deactivated olefins reactive toward the B₁₂-based photoinitiators, enhancing the photooxidizing power of the hypercloso-cluster was required. Since the redox potential of perfunctionalized clusters can be tuned by varying the substituents bound to the cluster, more electron-withdrawing C₆F₅ groups were installed to form hypercloso-B₁₂(OCH₂C₆F₅)₁₂. The photooxidizing power was sufficiently increased to enable the activation of a large variety of olefins, including isobutylene, a notoriously difficult monomer that typically requires metal-based activators or harsh conditions (Fig. 9D).⁹⁸ Most impressively, these reactions were observed to proceed with as little as 0.005 mol % initiator loadings.

Figure 9.

A. TD-DFT calculation describing the charge-transfer excitation pathway in hypercloso-B₁₂(OCH₂C₆F₅)₁₂. B. Schematic representation of hypothesized docking of styrene monomers among the pentafluoroaryl groups of hypercloso-B₁₂(OCH₂C₆F₅)₁₂ which helps facilitate electron transfer. C. UV-vis absorption spectra for neutral hypercloso-B₁₂(OCH₂C₆F₅)₁₂ (black) and radical hypocloso-[B₁₂(OCH₂C₆F₅)₁₂]¹⁻ (red). D. Production of branched poly(isobutylene) initiated by B₁₂(OCH₂C₆F₅)₁₂ in the presence of blue LED light.

Applications as Weakly Coordinating Ions

Catalysis

Ozerov and co-workers have leveraged the weakly coordinating nature of closo-[B₁₂X₁₂]²⁻ dianions in the catalytic hydrodefluorination (HDF) of sp³ C–F bonds in the presence of silylium cations generated in-situ from H–SiEt₃.⁶⁵ In dichlorobenzene solvent, >97% HDF of 4-fluorobenzotrifluoride occurred at room temperature in 1h; the more challenging substrate, perfluorotoluene, likewise underwent HDF in 1h at 80°C. In both cases, perfect selectivity was observed for sp³ C–F bonds and arene C–F bonds remained intact. This method is a clear proof-of-concept that perfunctionalized dodecaborates can be successfully employed as weakly coordinating anions under catalytic conditions. Furthermore, this example shows that these dianions can behave as competent substitutes for carba-closo-dodecaborate analogues (CB₁₁H₁₁¹⁻ and functionalized derivatives) which, despite their well-established utility, are more difficult to synthesize, more expensive to purchase, and generally more difficult to perfunctionalize.

More recently, Kirsch and co-workers successfully employed a perfunctionalized boron cluster under gold catalysis conditions to effect a variety of ene-yne cyclization reactions. Single crystal X-ray structural data was also obtained which revealed a dimeric active catalyst and the weakly coordinating nature of the dodecaborate counterion.⁹⁹ The kinetic stability of these perfuctionalized boron clusters is particularly relevant to (cationic) transition metal catalysis, since even the widely used BAr^F anion ([B(C₆F₅)₄]⁻) has been shown to engage in aryl ring transfer in the presence of transition metal cations.^100,101,102

Reagents

The weakly coordinating nature of these dianions prompted Knapp and co-workers to employ [CH₃]₂[B₁₂Cl₁₂] as a methylating agent. Me⁺–Cl interactions stabilize the methyl cation however the dodecaborate framework remains intact. This compound is capable of methylating SO₂ as well as benzene to form [CH₃OSO]₂[B₁₂Cl₁₂] and [CH₃C₆H₆]₂[B₁₂Cl₁₂], respectively.¹⁰³ The same group later disclosed the stabilization of [Al(CH₃)₂]⁺ using this dodecaborate anion.¹⁰⁴ It should be noted that Wehmschulte and co-workers reported a similar complexes employing a perfunctionalized dodecaborate as a counterion, including [Ag(CH₃CN)][B₁₂Cl₁₁(NMe₃)], [CPh₃][B₁₂Cl₁₁(NMe₃)], and [Et₂Al][B₁₂Cl₁₁(NMe₃)].¹⁰⁵

Superacids

Superacids have been used for many years to probe extremes of chemical behavior and engage interesting reactivity. Among the strongest are those based on carba-closo-dodecaborates, which like their neutral carborane and dianionic dodecaborate analogues exhibit chemical stability toward acidic solutions as well as charge delocalization which is thought to enhance acidic properties. And while most attention has been given to CB₁₁-based superacids, comparatively fewer studies have examined the acidity of dianionic dodecaborate derivatives. Muetterties and co-workers noted the extreme acidity of B₁₂Cl₁₂ and B₁₂Br₁₂ in the mid-1960s in which the authors disclosed their superior acidity to sulfuric acid.⁵³ More recently, experimental^106,107 and theoretical^107,108 efforts have been made to further quantify the (potential) acidity of such derivatives. In particular, Reed and co-workers showed¹⁰⁶ that B₁₂Cl₁₂ and B₁₂Br₁₂ exhibit comparable acidity to carborane acids and are sufficiently acidic to protonate benzene, which even trifluoromethanesulfonic acid and fluorosulfonic acid cannot do. Furthermore, both protons from H₂B₁₂X₁₂ are capable of arene protonation, not simply the first. This is a fundamental limitation of carborane acids and suggests potential for applications where superacidic conditions are necessary.

Ionic Liquids

Ionic liquids are gaining increased attention due to their unique chemical properties and potential applications.¹⁰⁹ Polyhedral boron clusters stand as viable candidates for such an application due to their known chemical and electrochemical stability as well as their tunability (vide supra). While usually encountered as solids, recent work has shown that these species may be appropriately functionalized to give low-melting (< 100°C) ionic liquids. Jenne and Kirsch reported that perfunctionalized clusters of the type [C₆mim]₂[B₁₂X₁₁OR] ([C₆mim]⁺ = 1-hexyl-3-imidazolium) displayed decomposition temperatures well beyond 300°C for five of six derivatives. One of these, [C₆mim]₂[B₁₂Cl₁₁(O-ⁿPr)] gave a melting point under 100°C.⁸³ More recently, the same group showed improved melting properties of dodecaborate-based ionic liquids using [B₁₂X₆H₅NR₂]²⁻ and [B₁₂X₆H₅NR₃]⁻ anions.¹¹⁰ These compounds show low (as low as 57°C) melting points and high (> 2V) electrochemical stability, showing promise not only as ionic liquids but also as electrolytes. Notably, [C₆mim][B₁₂Cl₆H₅N(ⁿPr)₃] shows a liquid range of over 300°C between melting (65°C) and decomposition (371°C).

Materials with cationic conductivity

In early works on Li-ion batteries, several research groups evaluated anionic boron clusters including closo-[B₁₂Cl₁₂]²⁻ and closo-[B₁₀Cl₁₀]²⁻ as electrolyte components.¹¹¹ For example, primary liquid cathode Li/SOCl₂ cells, which utilized Li₂[B₁₀Cl₁₀] and Li₂[B₁₂Cl₁₂] salts were reported in 1979.¹¹² Later, these electrolytes were tested in Li/TiS₂ cells by Exxon.^113,114 More recently, fluorinated all boron nanoclusters Li₂[B₁₀F₁₀] and Li₂[B₁₂F₁₂] salts were developed at Air Products (commercial name – “Stabilife fluorinated electrolyte salts”). The electrolyte solvent/salt formulation can be tuned to obtain promising cell cycling performance, especially at elevated temperature (60°C).^115,116 The high electrochemical stability of these anions has been shown to provide overcharge protection for 4V Li-ion batteries, shown with MCMB/spinel cells. Perfunctionalized cluster anions containing protons as counterions have been investigated for their potential use in proton conductor membranes. Early efforts testing H₂[B₁₂Cl₁₂] however proved unsatisfactory for this application given significant decomposition of this cluster observed at elevated temperatures in aqueous media resulted from B-Cl bond hydrolysis.⁹⁰ In 2004, Stasko and co-workers reported the solid-state proton conductivity study of closo-H₂[B₁₂(OH)₁₂], the conjugate acid of the B₁₂(OH)₁₂ dianion (Fig. 10).¹¹⁷ The title compound was generated by treating closo-Cs₂[B₁₂(OH)₁₂] with concentrated HCl at 150°C for 3d at ~8.5 atm. As with other dodecaborate based anions, closo-H₂[B₁₂(OH)₁₂] displayed excellent thermal stability (>400°C in air or N₂ atmosphere). A proton conductivity of 1.5 × 10⁻⁵ Ω⁻¹ cm⁻¹ was measured for H₂[B₁₂(OH)₁₂], which is an order of magnitude higher than that of known solid-state proton conductor CsHSO₄. This conductivity was shown to increase with increasing temperature. Distinct changes in proton dynamics at varying temperatures were also observed by solid-state ¹H NMR, corroborating the proton mobility hypothesis.

Figure 10.

Schematic representation of the crystal lattice of closo-H₂[B₁₂(OH)₁₂] synthesized by Stasko and co-workers which was shown to conduct protons in the solid state.

Very recently, Strauss and co-workers disclosed detailed studies of the coordination of Na₂[B₁₂X₁₂] (X = F, Cl) and their hydrated forms to Na⁺ ions in the solid state.¹¹⁸ Interestingly, they concluded that B₁₂Cl₁₂²⁻ coordinates Na⁺ less strongly than B₁₂F₁₂²⁻ and ultimately suggest that both perfunctionalized dodecaborate derivatives, in their dehydrated forms, may be viable candidates for ion conductivity in energy storage applications.

Biological Applications

A number of boron cluster derivatives are attractive candidates for a broad range of biomedical applications due to their exceptional stability and biocompatibility, availability of ¹⁰B nuclei (20% natural abundance), and tunability.^52,119–121 Over the past few decades, many biological applications of boron clusters have been explored, most notably in the areas of drug delivery and imaging for the diagnosis and treatment of cancer,^7,119 and only most recently in the area of probing protein-biomolecule interactions.⁴⁵

Delivery of ¹⁰B for BNCT

The most widely investigated research area within boron cluster-based drug delivery is undoubtedly boron neutron capture therapy (BNCT), a binary therapeutic strategy for the treatment of cancer (Fig. 11A). The concept of BNCT was first proposed in 1936 by Locher,¹²² and has since captured the attention of many researchers due to its elegant design. As described by Hawthorne and others, the beauty of BNCT lies in the fact that both of its essential components, nonradioactive ¹⁰B nuclei and low-energy thermal neutrons, are nontoxic by themselves, however together they initiate an energetic nuclear decomposition reaction that typically cannot extend past the diameter of a cell.^10,119,123 Based on this principle, BNCT was proposed as a safer alternative to many of the single-component radiation and chemotherapy approaches for the treatment of cancer. Many boron-containing compounds have been investigated for BNCT with generally unsuccessful results, with the exception of L-p-dihydroxyborylphenylalanine (BPA) (Fig. 11B), which is still being used in the clinical settings today.^119,124,125 Between 1959 and 1960, the discovery of polyhedral borane anions closo-[B₁₀H₁₀]²⁻ and closo-[B₁₂H₁₂]^{2− 5,6,126} sparked renewed interest in BNCT given the exceptional stability of these species and their high boron content. The sodium salt of [B₁₀H₁₀]²⁻ (Fig. 11B) was tested in animals, and while it showed promising results, it later failed in a clinical trial.^127,128 Two thiol-containing derivatives, Na₂[B₁₀Cl₈(SH)₂] and Na₂[B₁₂H₁₁SH] (BSH) (Fig. 11B), were identified as potential drug candidates from studies using mouse tumor models.^128,129 Although Na₂[B₁₀Cl₈(SH)₂] proved too toxic in mice, BSH exhibited more favorable properties and has since been extensively used in the clinical trials.^96,121,130

Figure 11.

A. An illustration of BNCT, a binary strategy for the treatment of cancer. B. Some BNCT candidates used in the preclinical and clinical settings.

However, despite the relative success of BPA and BSH in the clinical trials, an important challenge remains to selectively deliver sufficiently large amounts of ¹⁰B nuclei to the tumor tissue (approximately 20–35 μg/g tumor or 10⁹ atoms/cell), which is critical to the success of BNCT.^10,131 Therefore, two general approaches have been taken to tackle this issue. The first strategy involves the incorporation of boron into biomolecules with tumor-targeting abilities such as peptides and proteins, nucleic acids, carbohydrates, and porphyrins,^132,133 while the second approach uses liposomes and nanoparticles to deliver ¹⁰B atoms to tumor tissues.^133–137 In particular, many boron-encapsulated and boron-lipid liposomes have shown promising results in their ability to selectively deliver therapeutic quantities of ¹⁰B atoms into tumor cells for effective BNCT.^138–146

However, while the liposome-based approach has been successful, this system is inherently inefficient, as only around 5–10 % of the total mass arises from boron.^73,147 In an attempt to improve the efficiency, boron cluster-based nanoparticles have been proposed as alternative BNCT agents. Toward this end, Hawthorne and colleagues synthesized ester-linked⁷³ and ether-linked¹⁴⁷ derivatives of the dodecaborate core with pendant nido-o-carboranyl substituents, which in total contain roughly 40% boron content by mass. The sodium salt of the ether-linked cluster (Fig. 12) was found to be water-soluble and was subsequently evaluated for its suitability as a BNCT agent in a series of toxicity and biodistribution studies in animals.¹⁴⁷ The results from these studies indicate that the compound is non-toxic to mice at a therapeutic dose and tumor-selective (the tumor-to-blood ratio is 9.4), however in this particular study its accumulation in tumor tissues (10.5 μg/g tumor) was half of the therapeutic amount necessary for BNCT. Nevertheless, this strategy of using cluster-based nanoparticles as BNCT agents is appealing due to its high boron content, modularity, and low toxicity.

Figure 12.

A. The chemical structure and B. A 3D representation of the dodeca(nido-o-carboranyl)-substituted dodecaborate cluster.

Very recently, Kovalska, Gumienna-Kontecka, and co-workers reported a spectroscopic and calorimetric binding study of perhalogenated decaborate ([B₁₀X₁₀]²⁻, X = Cl, Br, I) and dodecaborate ([B₁₂X₁₂]²⁻, X = Cl, I) dianions to bovine and human serum albumins (BSA and HSA, respectively) toward potential use in BNCT.¹⁴⁸ It was shown through isothermal titration calorimetry (ITC) and fluorescence quenching experiments that the halogenated boron clusters displayed a greater affinity toward albumin binding (~ 4–5 clusters per protein, K ~ 10⁴ – 10⁶ M⁻¹) than the parent, [B_nH_n]²⁻ (n = 10 or 12) cluster (2 clusters per protein, K ~ 10³ M⁻¹). Circular dichroism studies suggested that no significant change in the secondary protein structure had occurred even at the highest concentrations of cluster species tested, which is consistent with the inefficient fluorescence quenching of the HSA, which contains a Tyr residue buried within the protein manifold, relative to BSA, which contains both buried and surface Tyr residues, by all clusters examined. In a similar study, Losytskyy et al. studied the binding of oligoether- and ammonium-substituted decaborate anions to BSA, HSA, immunoglobin IgG, b-lactoglobulin, and lysozyme. It was found that the functionalized, charge-compensated B₁₀-based monoanions formed stronger interactions with these protein substrates than with unfunctionalized decaborate and dodecaborate dianions.¹⁴⁹ These studies present an additional delivery system for high boron content molecules in vitro and suggest that perhalogenated boron clusters may be potentially useful in BNCT applications.

Delivery of Chemotherapy Drugs

In addition to their function as BNCT agents, boron clusters have recently been developed as nanocarriers for the delivery of various chemotherapy drugs. Within the past few years, Hawthorne and colleagues have constructed several heterofunctional cluster-based drug delivery systems. For example, Bondarev et al. demonstrated the synthesis of a vertex-differentiated dodecaborate cluster scaffold, closo-[B₁₂(OH)₁₁NH₃]^1–, that is further functionalized with an amide-bound fluorescein and eleven carboplatin analogues bound through carbamate linkages.⁸⁴ This species is found to localize in the nuclei of A549 cancer cells and shows potent cytotoxicity against both platinum-sensitive and platinum-resistant cell lines.⁸⁴ In another example by Sarma et al., a trimodal targeted drug delivery system was assembled from a hetero-perfunctionalized dodecaborate dianion, closo-[B₁₂(OH)₁₁)(OR)]^2–, by conjugating one vertex to a fluorescein molecule via click chemistry and the eleven other vertices to a branched linker featuring both the targeting molecule glucosamine and the anticancer drug chlorambucil (Fig. 13A).¹⁵⁰ Interestingly, the targeted cluster co-localized with the lysosomes (Fig. 13B) and mitochondria within Jurkat cancer cells and exhibited 15-fold enhanced cytotoxicity compared to the non-targeted cluster against cells overexpressing the receptor for glucosamine.¹⁵⁰ In a different approach to assembling the cluster-based nanocarriers, encapsulation of hydrophobic chemotherapy agent doxorubicin by the dendritic dodecaborate clusters in a manner reminiscent to its encapsulation by the poly(amidoamine) (PAMAM) dendrimers was disclosed,^151,152 thus laying the groundwork for potential delivery into cancer cells using funtionalized dodecaborates.

Figure 13.

A. A trimodal closomer drug delivery system, CDDS-1, features targeting, imaging, and cell-killing abilities. B. Confocal microscopy images of CDDS-1 show its co-localization with lysosomes. Reprinted (adapted) with permission from Sarma, S. J.; Khan, A. A.; Goswami, L. N.; Jalisatgi, S. S.; Hawthorne, M. F. Chem. Eur. J. 2016, 22, 12715. Copyright 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.

Diagnostic Imaging

In addition to their applications in drug delivery for various types of cancer treatments, many derivatives of boron clusters have also been investigated for potential applications in diagnostic imaging. In fact, for decades, radiolabeled boron clusters have been studied as X-ray contrast agents because the incorporation of radiolabels on boron clusters can lead to higher chemical stability of the radiolabels.⁷ However, while there have been numerous reports on the radiolabeling of boron clusters,^7,13,153 the vast majority have focused on the radiolabeled carborane derivatives and therefore will not be discussed here.

We will instead examine the development of high-performance all-boron cluster-based contrast agents for a more modern imaging technique, magnetic resonance imaging (MRI). MRI has emerged as one of the most important tools for the diagnosis of diseases, however there is a need to develop larger MRI contrast agents that can overcome the issues of low relaxivity values, short plasma half-lives, and low contrast typically associated with low-molecular weight contrast agents, while remaining well-defined.^154–157 To tackle this challenge, Hawthorne and colleagues reported the synthesis of both the non-targeted¹⁵⁴ and targeted¹⁵⁷ (Fig. 14A) dodecaborate cluster-based MRI contrast agents that feature multiple gadolinium(III) complexes, and further demonstrated that these compounds provided significantly enhanced image contrast in addition to tumor-targeting ability provided by the cyclic RGD (cRGD) peptide (Fig. 14B).

Figure 14.

A. The cRGD peptide-labeled, Gd³⁺-DOTA chelates-carrying dodecaborate cluster MRI contrast agent and B. the resulting strong contrast enhancement in T1-weighted MRI scans of mice for up to 4 hours after injection. Reprinted (adapted) with permission from Goswami, L. N.; Ma, L.; Cai, Q.; Sarma, S. J.; Jalisatgi, S. S.; Hawthorne, M. F. Inorg. Chem. 2013, 52, 1701. Copyright 2013 American Chemical Society.

Multivalent Protein Binders

We recently reported a strategy to rapidly assemble atomically precise, highly tunable organomimetic cluster nanomolecules (OCNs) via perfluoroaryl-thiol S_NAr chemistry (Fig. 15).⁴⁵ This approach is reminiscent of the facile assembly method of the thiol-capped gold nanoparticles (AuNPs), however the resulting OCNs are monodisperse and structurally robust under a range of biologically relevant conditions due to covalently-linked substituents on the cluster core.

Figure 15.

A. The pentafluoroaryl-based dodecaborate clusters can undergo efficient S_NAr reactions with a variety of thiolated molecules. B–C. Single X-ray crystal structures of hypercloso-B₁₂(OCH₂C₆F₅)₁₂ (B) and hypercloso-B₁₂(OCH₂-4-(C₆F₅)-C₆H₄)₁₂ (C).

Employing this strategy, the pentafluoroarylated dodecaborate clusters were perfunctionalized with a broad scope of thiolated (macro)molecules, including unprotected peptides, saccharides, and polymers. Furthermore, a glycosylated OCN was shown to be capable of multivalent interactions with Concanavalin A (Con A) (Fig. 16), resulting in binding affinities that are orders of magnitude higher than those observed for a single carbohydrate-lectin interaction. Taken together, this work showcases how one might rationally design and synthesize well-defined nanomolecules that can be useful in 1) elucidating multivalent binding mechanisms and 2) potentially probing or inhibiting protein-biomolecule interactions.

Figure 16.

A snapshot after 20 ns of a molecular dynamics simulation between a glycosylated OCN and the tetrameric lectin ConA. The carbohydrate-binding sites of ConA are highlighted.

Outlook and Conclusions

Beyond the applications mentioned above, perfunctionalized boron clusters remain relatively unexplored entities. Yet, these molecules represent a very unique class of building blocks for potential applications in materials science. For example, polymeric materials produced from boron-rich clusters have been disclosed as early as 1968,¹⁵⁸ but surprisingly, studies focused on producing functional, polymer-based materials featuring perfunctionalized boron cluster species have been extremely scarce. This is in contrast to a number of studies addressing the use of various carborane-based scaffolds for applications ranging from coatings to porous materials.^159,160 Thus, polymeric materials and surfaces featuring some of the halogenated clusters can expand the scope of existing materials and provide resins and supports featuring bulk “weakly coordinating” properties reminiscent of zeolites and other “weakly-coordinating” supports.¹⁶¹ Similarly, polymers featuring anionic boron clusters as pendant functional groups should produce a unique class of polyelectrolytes. These materials can potentially show enhanced stabilities and tunable affinities toward cations, which in turn should allow one to tune ion mobility in the solid state providing access to materials with rationally tuned compositions and morphologies.¹⁶² Furthermore, given the inherent three-dimensionality of substituted clusters, these should possess attractive properties as cross-linkers for existing classes of polymeric materials.¹⁶³

Clusters exhibiting pseudo-metallic properties manifested in their photoredox properties also have many unexplored avenues. First, understanding how functionalization affects the redox chemistry and the radical states of these highly delocalized systems is fundamentally important. The physical properties associated with unpaired electrons of functionalized species need to be studied further, both in the solid and solution states. These species can provide an entry to a unique chemistry of stable radicals, which can be used as building blocks with strong magnetic coupling properties to design molecular magnets.¹⁶⁴ Furthermore, these molecules can potentially exhibit useful properties for spin labeling and nuclear polarization transfer.^165,166 Given the potential materials-favorable properties of many perfunctionalized redox-active boron clusters, we expect that these species can be used as charge carrier scaffolds with good stability, cyclability, and potentially enhanced voltage characteristics in devices including polymer-based and redox-flow batteries.^167,168 Likewise, the ability to control geometry and electronic delocalization between the cluster and tethered substituents should allow one to tune charge-transfer properties of the resulting systems and ultimately design metal-free chromophores with tunable absorption/emission colors and excited state lifetimes.

Perfunctionalized boron clusters can also be envisioned as building blocks for hierarchical hybrid materials. Recent developments in the field of hybrid porous solids¹⁶⁹ clearly showcase the need for rigid, tunable, and robust three-dimensional building blocks. Furthermore, atomically precise 3D hierarchical materials containing nano-sized building blocks can be also used for assembly of nanoparticles into hierarchical extended structures. For example, cluster-based molecules can be synthesized with a densely packed corona of functional groups that will bind to the surface of nanoparticles. Tuning the aspect ratios of these clusters and nanoparticles may provide a method for creating assemblies with tunable plasmonic¹⁷⁰ and catalytic properties.¹⁷¹

Finally, in contrast to carbon-based aromatic building blocks where there exist numerous chemical transformations enabling the functionalization of C–H bonds, methods targeting heterofunctionalized boron clusters are still currently very limited. Metal-catalyzed transformations will likely emerge in the future targeting synthesis of these molecules.¹⁷² Significantly, many opportunities can be envisioned that stem directly from the ability to generate complex structures that contain multiple functional groups appended to its vertices (vide supra). Beyond metal-based catalysis, several studies indicate that perhalogenated carboranes and boron clusters can undergo substitution at B–Cl and B–F vertices,^173,174 suggesting it should be possible to design more well-behaved reaction protocols for creating vertex-differentiated systems via this relatively unexplored pathway. We are strongly convinced that expanding a molecular toolbox of existing organic and hybrid monomers for polymer synthesis through introduction of new perfunctionalized boron-rich cluster architectures will create a fundamentally new class molecular materials.