Boron Cluster Organocatalysis

Homogenous catalysis is a ubiquitous tool in the hands of practicing chemists allowing to prepare many molecules and materials ranging from commodity chemicals to pharmaceutically active ingredients.¹ One of the most abundant classes of homogeneous catalysts are transition metal-containing compounds. One great appeal of these compounds as catalysts is their modularity and tunability that is normally achieved through the use of ligands which coordinate to the catalytically-active metal site and provide electronic and steric modulation necessary for productive catalytic reactivity. However, metal-based coordination compounds are not the only ones that can engender catalysis. Over the past several decades, organocatalytic systems,² which do not consist of any metals, have grown as a separate class of powerful catalysts for many important chemical transformations. Interestingly, common transition metal ligands such as phosphines, N-heterocyclic carbenes, and organosulfur compounds have also been found to act in a catalytic manner for numerous reactions.² In much the same way that a metal center’s electronic and steric properties can be tuned by ligands, the electronic and steric properties of the reactive non-metal center (e.g., C, P, S) of an organocatalyst can be tuned by classical alkyl and aryl substituents (Figure 1A).³ The use of alkyl and aryl substituents, and a mix thereof, can impart differing electronic and steric effects on the catalytically active site of the organocatalyst, and further creative use of redox active moieties and chiral substituents can impart greater tunability of these systems (Figure 1A). Despite the breadth of alkyl and aryl substituents available to tune organocatalyst systems, these functional groups provide only a finite range of desired control. Therefore, researchers have been working on identifying nonclassical substituents to expand this range that could potentially lead to new paradigms in catalysis. In this issue of Chem, Kona and co-workers showcase an enticing new opportunity for catalysis, where authors show how “organomimetic” icosahedral carboranes can serve as a tunable substituent on a heteroatom ultimately delivering an unprecedented catalytic performance.⁴

Figure 1.

A. Representative organocatalytic platforms and substituent strategies used to tune these. B. Carboranes and their key properties. C. Development of the carborane halogenation catalysts as described by Kona and co-workers.⁴

Icosahedral carboranes (C₂B₁₀H₁₂) represent a broader class of polyhedral borane clusters that exhibit a unique electronic characteristic known as three-dimensional aromaticity due to the delocalized bonding nature of the polyhedral cage.⁵ Due to this three-dimensional aromaticity, carboranes exhibit electronic characteristics similar to that of two-dimensional aromatic species whilst maintaining a three-dimensional molecular footprint with a steric size akin to adamantane. Icosohedral carboranes can exist in three isomeric forms dependent on the position of the two carbon-based vertices present within the cage: the ortho, meta and para isomers (Figure 1B). In the case of the ortho and meta isomers, the positioning of the carbon atoms in the cage induces a dipole across the cluster due to the electronegativity difference between carbon and boron atoms (Figure 1B). This induced dipole is of specific interest as it greatly changes the electronic effects enabled by carboranes when these are bound to a functional group at a specific vertex. For example, it has been previously observed that heteroatoms bound to the carbon vertex of a carborane cage experience a powerful electron withdrawing effect larger than the one induced by perfluoroaryl groups.⁶ On the other hand, heteroatoms bound to boron vertices antipodal to the carbons (B9 and B10 vertices) experience a stronger electron donating effect compared to the classical alkyl-based substituents.⁷ These vertex dependent substituent effects have previously been leveraged in ligand design producing multiple classes of heteroatom substituted systems with extreme properties including extremely electron rich phosphines, and super Lewis acidic trisubstituted boranes (Figure 1B).^8,9 From the operational handling standpoint, carboranes behave very similarly to classical organic molecules in terms of high solubility in conventional non-polar solvents, high thermal and chemical stability and existence of a variety of straightforward methods available to derivatize these compounds. These features therefore render these compounds effectively “organic” (or to be more precise - “organomimetic”),⁷ even though they might not appear as such to an uninformed practitioner.

Kona and co-workers have previously reported that a thioether based catalyst containing a triptcenyl substitutent on the sulfur (Trip-SMe) can mediate electrophilic aromatic halogenation.¹⁰ However, this system presents limitations in terms of substrate scope and tunability. In their report, authors leveraged the use of carboranes as tunable substituents in the development of a new class of thioether-based organocatalysts that surpass previous limitations. The designed organomimetic thiomethyl-carborane (SMe-carborane) catalyst was shown to activate N-bromo-succinimide (NBS) in the presence of AgSbF₆ co-catalyst in order to brominate 2,4-dichloroanisole under mild conditions with low catalyst loadings (0.2–2%). Initially, authors tested multiple SMe-carborane catalysts with various substitution patterns on both boron- and carbon-based vertices in ortho- and meta-carborane (Figure 1C). Importantly, when the SMe- substituent was placed on the carbon vertex of meta-carborane, the catalytic activity of the resulting cluster showed the greatest promise. The system could be further fine-tuned through functionalization at the antipodal boron vertices. Thus, the most efficient catalyst in the series, termed “Cat H”, was developed featuring a B-octomethyl substituted meta-carborane with two SMe-based substituents bound to the carbon vertices (Figure 1C).

Further optimization of this reaction using the Cat H and AgSbF₆ system showed rapid conversion of many activated arenes at room temperature, however, less activated substrates such as clofibrate required a mild heating (40°−80° C). Impressively, the reaction reported can proceed as fast as one minute at room temperature as demonstrated in the case of monobromonation of Diclofenac™ (Figure 1C). In general, Cat H was shown to operate well with halobenzene substrates resulting in the selective formation of para-substituted products. In addition, this catalytic reaction tolerates numerous functional groups including olefinic substrates. Further substrate scope studies identified optimized conditions for the conversion of moderately deactivated arenes as well as molecules bearing other reactive substituents such as diselenides and alkynes. For more deactivated arenes, the use of Cat E (Figure 1C) and MeNO₂ solvent was implemented in order to improve the reaction conversion.

Late-stage modification of bioactive molecules is of great utility and access to halogenated species ultimately allows to utilize these molecules for further modification via various metal-catalyzed cross-coupling methods. Kona and co-workers illustrated this application using fluorenone modified via their method as a representative example where authors were able to arylate, alkynylate, and cyanate 2-bromo-fluorenone through palladium cross-coupling (Figure 1C). From the reactivity standpoint, the reaction is also general in terms of the halogen source, where N-chloro-succinimide and N-iodo-succinimide can be used successfully with Cat H in halogenating complex functional molecules.

To support their mechanistic hypothesis, a sulfonium salt meant to represent the catalytic intermediate was synthesized and isolated (Figure 1C). This compound can be reacted in catalytic amounts with 2,4-dichloranisole and N-bromo-succinimide to yield the brominated aryl product, suggesting that the isolated intermediate is likely catalytically relevant. Based on these and other observations, the authors propose a catalytic cycle invoking first the nucleophilic abstraction of halogen from N-halo-succinimide forming a sulfonium salt. This reactive high-valent sulfur intermediate is stabilized by the carborane cluster and is capable of engaging in electrophilic aromatic substitution chemistry. While under the standard catalytic conditions that the authors developed, Ag(I) salt is used as an additive to facilitate the process, it is not ultimately required as authors clearly demonstrate in control experiments where the isolated sulfonium can be used as a competent catalyst on its own.

The multiple SMe-carborane catalysts developed, yielded varying reactivity based on the vertex substitution and functionalization of the carborane substituent in relation to the reactive SMe group. Due to the similar steric profiles of the carborane substituents, electronic effects were determined as the key influencing parameter behind the catalyst’s differing reactivity. Two electronic parameters of interest were highlighted based on the proposed catalytic mechanism: the nucleophilicity of the sulfur and the energy level of the molecular orbital corresponding to the σ-hole of the sulfonium-based catalytic intermediate. Nucleophilicity of sulfur governs halide abstraction from N-halo-succinimide, while the energy of S-Br σ* orbital is responsible for the halogen transfer step. A computational study of these parameters in reference to the synthesized carborane catalysts as well as Trip-SMe illustrates a range of nucleophilicity and σ-hole energy levels. A delicate interplay of the above-mentioned parameters in Cat H specifically places this molecule in a sweet spot in terms of catalytic reactivity. Importantly, authors show that solvent effects can play a significant role in the catalyst development (likely due to ion pairing) and one can circumvent this by the use of Cat E, which unlike Cat H, can operate in MeNO₂ as necessitated by some substrates.

The work by Kona and co-workers highlights how non-classical substituents such as carboranes can compete with existing state-of-the-art organic-based tools to produce reactive and selective organocatalysts. Carboranes, depending on their specific attachment, can exhibit dramatically varied electronic effects when connected to a SMe group ultimately providing a highly tunable scaffold for organocatalyst design. It could be envisioned that other main-group organocatalytic systems would benefit from non-classical substituents such as boron clusters to fine tune the electronics of the reactive sites in a way that conventional aryl and alkyl substitution cannot.