Metal-mediated Coupling of Carbon Dioxide and Aryl/Vinyl Chlorides under Ambient Conditions

The field of carbon dioxide chemistry is undoubtedly among the fastest growing areas of chemical research with CO₂ representing an alternative, renewable, and cheap carbon feedstock for our fossil-fuel-based industry and economy.1 The key challenge using CO₂ as a reagent is its high stability, and in this context catalysis plays an extremely important role to lower the energy barriers associated with its high kinetic stability. The CO₂ “fixation” arena has witnessed an impressive growth in possibilities for organic chemists to convert this C 1 building block in many different and important products including organic (poly)carbonates,2 salicylic acid,3 urea,4 and chemicals obtained by means of the reductive conversion of CO₂.5 Despite the progress made, the field still faces important challenges to be solved, such as the development of catalytic procedures that can meet the criteria of sustainable production, while maintaining a high level of reactivity and selectivity. In this respect, (transition) metal catalysis may offer a versatile strategy towards the activation of substrates under particularly mild (i.e., ambient) reaction conditions.6

Catalytic (hydro)carboxylation (Scheme 1)7 employing CO₂ has become a popular and viable approach for the synthesis of a variety of carboxylic acid synthons useful in both academic and industrial settings.8 Although the carboxylation of heterocyclic substrates and alkynes is relatively easy, owing to the higher pKa values of the involved CH fragments,9 the catalytic approach towards the conversion of aryl-H10 or more reactive aryl-X (X=halide) bonds11 into carboxylated functions by using CO₂ represents a much greater challenge. In this respect, the group of Martín has achieved a significant step forward by using a direct carboxylation approach through a Pd-catalyzed reductive carboxylation of various aryl bromides. This procedure combines the use of a bulky phosphine (i.e., 2-di-tert-butylphosphino-2′,4′,6′-tris-iso-propyl-1,1′-biphenyl) and Pd(OAc)₂, and provides an interesting and extensive substrate scope that allows conversion of aryl bromides comprising of various other functionalities, such as amines, aryl chlorides, esters, oxiranes, thiophenes, and even tolerates ortho-substituents. Though seminal in nature, the Martín protocol shows some limitations: vinyl halides and aryl chlorides (which are cheaper than their corresponding bromides) proved to be unreactive, air-sensitive and pyrophoric diethyl zinc was used as a reducing agent and in some cases relatively large amounts of by-product were formed. Inspired by this work, Tsuji and co-workers12 have now developed a simple catalyst system based on readily available components and additives that can largely address these issues. These latter advances create therefore a new and important stimulus for the transition metal catalyzed conversion of carbon dioxide, and provide a new dogma for synthetic chemists in their quest to efficiently use this renewable building block for improving and expanding its product scope. It may also be regarded as an encouraging starting point for exploring new and interesting reactivity profiles of new catalyst systems13 on their way to realize maximum sustainability. Because of their importance, the results of Tsuji are detailed and highlighted herein.

Scheme 1. The reductive carboxylation of aryl halides and pseudo-halides using CO₂ as reagent.

The catalyst system developed by Tsuji (Scheme 2) comprises of a simple nickel source [NiCl₂(PPh₃)₂] using Mn powder (99.99 % purity) as reducing agent and Et₄NI as additive; 1,3-dimethyl-2-imidazolidinone (DMI), an aprotic cyclic urea, was used as medium in the screening stage with 1-butyl-4-chlorobenzene as substrate at 25 °C and only 1 atm of CO₂ pressure. These latter conditions are also significantly better than those reported in the state-of-the-art,11 and the catalyst structure is cheap and accessible. In the absence of Mn, Et₄NI, the Ni^II complex or CO₂ virtually no conversion of the aryl chloride was noted. A co-catalytic amount of PPh₃ (10 mol %) proved to beneficial as in the absence of it lower amounts of desired product (53 % by GC, analyzed as the methylester, Scheme 2) were formed and a relatively high amount of the by-product (biaryl, 27 %). Thus, all these additives were required for conversion of the aryl chloride into the carboxylated product. In addition, other phosphine ligands such as PCy₃, dppe, more electron-rich triarylphosphines or 2,2′-bipy resulted in significantly lower or no conversion of the substrate. The role of the Mn was also underlined by comparison of its effectiveness in this protocol with other reducing agents such as Zn and Mg, with the latter two ones being incapable of efficiently mediating this conversion. The role of the solvent was also investigated; DMI proved to be the preferred choice as comparative reactions carried out in DMF, THF or toluene furnished the desired product (Scheme 2) in much lower amounts.

Scheme 2. The catalytic system reported by Tsuji for the carboxylation of 1-butyl-4-chlorobenzene. The results shown here are related to the optimal conditions.

A wide range of aryl chlorides can be converted into their carboxylic acid derivatives using this newly developed protocol, among which are electron-rich and -poor substrates, and examples include also those containing ester or amide functions (Scheme 3). Other functionalities were also tolerated by this Ni-based catalyst, such as those containing boronic acid, thiophene, and functional polycyclic scaffolds. Although in the majority of the studies cases aryl chlorides were used, alternatively aryl tosylates and triflates can also be suitable substrate candidates. The limitation of this catalytic protocol was demonstrated by the attempt to convert ortho-substituted aryl chlorides or substrates that comprise of (unprotected) OH or NH₂ fragments; in these cases the procedure turned out to be ineffective. Remarkably, aliphatic and conjugated vinyl chlorides could also be converted into the desired carboxylates (Scheme 3, with the vinyl fragments highlighted in blue) in appreciable yields. Thus, these data further demonstrate the usefulness of this chemistry as providing the only known successful examples for this substrate category to date. To efficiently convert these vinyl chloride starting materials, the PPh₃ ligand in the Ni^II complex needed to be replaced by 2,2′-bipy, whereas the other additives and conditions remained unchanged.

Scheme 3. Selected examples of the substrate scope for the conversion of various aryl and vinyl chlorides using the catalyst system reported by Tsuji. Here, the products were isolated as the carboxylates unlike in Scheme 2. * This product was made from the corresponding aryl triflate.

Despite the demonstrated potential of Tsuji′s Ni^II-catalyst in the carboxylation of aryl/vinyl chlorides, there remain some challenges to overcome related to the substrate scope. Hence it is crucial to have a deeper understanding of the mechanism operative in these conversions. Preliminary work by the authors has revealed the likeliness of having a Ni^I intermediate (Scheme 4) that results from: 1) reduction of the Ni^II starting complex to form a Ni⁰ species (cf., Mn reducing agent), followed by 2) a subsequent oxidative addition of the aryl chloride (forming a Ni^II intermediate), and a 3) second reduction step of the latter to furnish a Ni^I species that, 4) in the presence of CO₂ gives the insertion product from which 5) the aryl-COO fragment reductively eliminates. The possible presence of a Ni^I species seems plausible as supported by electrochemical observations and previous results obtained in electrochemical carboxylations.14 What still remains elusive is the exact role of the additive Et₄NI (an investigation of the type of halide or counter-cation was not conducted) and how it is entangled in the mechanistic cycle. Furthermore, though not reported, there seems to be optimal activity when three equivalents of Mn are used though only two molar equivalents are required. Kinetic studies would be helpful to evaluate the proposed mechanistic pathway and to examine the order dependence. Other useful tools could include computational methods and EPR to elucidate the presence of this Ni^I intermediate.

Scheme 4. Proposed mechanistic cycle by Tsuji.

There are only a few methods available for the direct carboxylation of aryl halides that show expanded application potential.11, 12 The newly developed protocol by Tsuji undisputedly shows the great potential that transition metal catalysis can offer to realize more sustainable, more effective and more attractive solutions for challenging conversions. The potential of the Tsuji system is indeed impressive, though some minor weaknesses (e.g., in the substrate scope, moderate yield for some of the studied substrates, and the exact role of the additives in the mechanistic cycle) provide topics for further research and optimization. Nevertheless, the realization of a carboxylation process for the cheap aryl chlorides, using an attractively simple and accessible catalyst system based on a Ni^II complex, and employing an easy to handle reducing agent (Mn) under extremely mild conditions (25 °C, 1 atm CO₂) is a major step forward in the effective conversion of CO₂ into value-added commodities.

Acknowledgements

The author thanks ICIQ, ICREA, and the Spanish Ministerio de Economía y Competitividad (MINECO) through project CTQ2011-27385 for their support.