Carbon isotope exchange and transfer reactions involving CO₂

Abstract

Carboxylation reactions are fundamental transformations in organic chemistry, enabling the synthesis of diverse and valuable compounds, including polymers, pharmaceuticals, and agrochemicals. In recent years, carboxylation processes proceeding via carbon isotope exchange or CO₂ transfer have received significant interest. These processes offer intriguing decarboxylation/carboxylation reactivity and enhance existing carboxylation pathways, in particular by delivering efficient and sustainable solutions for carbon isotope labelling. This review compares and contrasts the related fields of carbon isotope exchange and transfer reactions involving CO₂. We detail early developments to the most recent advancements and discuss the advantages and limitations of each approach, in particular with regard to the efficiency of each strategy.

---

1. Introduction

Isotope chemistry represents a unique and essential field with broad relevance in both academic research and industrial applications. In particular, the pharmaceutical and agrochemical sectors rely heavily on isotopically labelled compounds.^1,2 More specifically, ¹¹C isotopes play a critical role in therapeutic research, as the high sensitivity enables high-resolution imaging and tracking using positron emission tomography (PET) for biodistribution, pharmacokinectic and receptor binding studies. The radioactivity facilitates the detection and mapping of various cancers.³ On the other hand, ¹³C labelling, in addition to applications in structural elucidation and mechanistic studies, has become a powerful bioanalytical tool for investigating metabolic pathways, including studies of cancer metabolism.⁴ Meanwhile, ¹⁴C-labelled compounds are extensively used in in vivo metabolic (ADME) and pharmacokinetic (DMPK) studies.⁵ The long half-life (t_1/2 = 5730 years) and low-energy β-decay provide a safer and higher-resolution alternative to other isotopes.⁶ Furthermore, isotopic labelling is also valuable in agrochemical development, enabling detailed metabolic profiling and assessment of potential human and environmental exposure.⁷

Carboxylation is a cornerstone transformation in organic chemistry, underpinning the preparation of various molecules, including polymers, pharmaceuticals, and agrochemicals.⁸ Although all labelled reagents are relatively costly compared to their non-labelled counterparts, labelled CO₂ gases are the least expensive (¹³CO₂, ¹⁴CO₂) or the easiest to prepare (¹¹CO₂).⁹

Thus, carboxylation technologies often present an ideal method for introducing carbon isotopes. However, some distinct considerations must be made when performing a process with labelled CO₂ in comparison to a standard carboxylation reaction (Scheme 1A), namely; (1) labelled CO₂ gas is significantly more expensive than non-labelled CO₂, thus, maximizing the yield relative to CO₂ is critical. (2) ¹⁴CO₂ is radioactive, therefore, methods using equimolar amounts of the labelled reagent are preferred to avoid waste. The development of non-gaseous CO₂ surrogates may also aid the handling of these hazardous reagents and obviate the need for specialised gas handling manifolds. (3) ¹¹CO₂ is a short-lived isotope (t_1/2 = 20 min), thus, short reaction times are required. Recent years have seen the emergence of carbon isotope exchange and transfer reactions with CO₂ as useful strategies that satisfy these considerations and will be the focus of this review (Scheme 1B and C). These processes have the potential to directly label complex carboxylic acids via single step synthetic routes, avoiding de novo synthesis and thereby minimising waste and handling steps. CO₂ transfer additionally offers bench-weighable surrogates and near-quantitative incorporation, providing a green and practical option for isotope labelling.

Scheme 1 (A) Comparison of labelled vs unlabelled CO₂. (B) Overview of carbon isotope exchange reactions. (C) Overview of CO₂ transfer reactions.

In carbon isotope exchange, the unlabelled carboxylic acid 1 functionality is swapped with labelled CO₂ gas to provide the labelled carboxylic acid 2 (Scheme 1B). A simplified description of this process involves decarboxylation of 1 to remove the unlabelled carboxyl group (step i), followed by carboxylation of the reactive intermediate 3¹⁰ with labelled CO₂ (step ii). Key advantages of this process include (1) switching one isotope for another presents a simple yet effective approach for installing isotope labels. (2) The process can be applied in late-stage isotope labelling. For example, complex molecules can theoretically be selectively labelled at the carboxylic acid group, whilst leaving the rest of the molecule untouched. This allows the molecule to be labelled directly and avoids lengthy de novo synthesis, thereby improving the overall greenness of the process. (3) If the carbon-isotope exchange is fast, it can be used for ¹¹C isotope labelling. A limitation of this method is that an excess of labelled CO₂ is required to deliver high levels of isotopic enrichment. This is because, upon decarboxylation, the intermediate 3 can react with the unlabelled CO₂ that has just been expelled (giving back the starting material 1) or react with the desired labelled CO₂. An excess of labelled CO₂ is therefore required to favour the desired isotope exchange, but this can add significant expense and waste. To improve the greenness of this process, the excess/unused labelled gas could be recycled, however, this is rarely demonstrated in the literature.¹¹ Lower equivalents of labelled CO₂ gas can be used, but at the expense of lower isotope incorporations. Although highly enriched isotopically labelled compounds are often not required, if the exchange is particularly inefficient, the isotope level may not meet the levels required for purpose. In this review, we focus our discussion on carbon isotope exchange reactions involving CO₂ to allow comparison with related CO₂ transfer reactions. For carbon isotope exchange more broadly, we direct the readers to several expert reviews.^9,12–14

The other strategy that will be discussed in this review is CO₂ transfer, sometimes referred to as CO₂ shuttling (Scheme 1C).¹⁵ A simplified description of this process involves the release of CO₂ from decarboxylation of a carboxylic acid-containing molecule 4 (step i), followed by the capture of the CO₂ by another species 5 via intermediate 8¹⁰ to deliver the desired carboxylated product 7 (step iii). As will be discussed in more detail below, the intermediate 3¹⁰ can also play a key role in the reaction, for example by acting as a base/metalating agent (step ii). We limit our discussion to CO₂ transfer reactions that are isodesmic, i.e. the types of bonds that are made in forming the products are the same as those which are broken in the reactants.¹⁶ More specifically, reagent 4 contains a C–CO₂ bond to deliver a C–CO₂ bond in product 7 (Scheme 1C). This allows us to better compare the related areas of carbon-isotope exchange and CO₂ transfer.¹⁷ CO₂ transfer presents several benefits in comparison to traditional carboxylation pathways including (1) the use of stable and weighable CO₂ surrogates. This improves the practicality of the procedure as it avoids the handling of gaseous reagents. This also presents safety benefits, for example by avoiding the use of high-pressure apparatus and the possibility of asphyxiating gas leaks. In addition, high vacuum transfers that are used when performing isotope labelling, including the liberation of ¹⁴CO₂ from labelled barium carbonate and subsequent condensation with liquid nitrogen, are avoided. (2) As the reagent can be weighed on the bench, it allows the equivalents of the labelled reagent to be easily controlled, thereby minimising costs and waste. (3) Due to inherent differences in the mechanisms, transfer reactions lead to higher (often full) isotope incorporation over carbon-isotope exchange. (4) We have largely described the benefits of CO₂ transfer reactions in the context of carbon isotope labelling, however, these methods can also find use in carboxylation reactions more generally as they provide practical and easy-to-handle CO₂ surrogates. Drawbacks include the need to prepare the carboxylating agent, which can add steps on to the synthesis whilst increasing costs and waste. Similarly, this can prevent applications in ¹¹C labelling due to the short lifetime of this isotope, however automated processes, may offer solutions to this problem.¹⁸ The atom economy of the process is also lower in comparison to using CO₂ gas, however, these drawbacks can be offset by practicality and cost benefits, especially when performing isotope labelling.

2. Carbon isotope exchange reactions

2.1. Early reports on carbon isotope exchange

One of the earliest examples of carbon isotope exchange via carboxylation dates back to 1974 when Szammer and co-workers postulated a process driven by thermal exchange for incorporating ¹⁴C isotopes (Scheme 2).¹⁹ Preliminary references to such transformations can be found in works led by Nakai and Darensbourg,^20,21 however Szammer provided the first detailed account and scope, specifically describing the process as isotope exchange. A range of labelled carboxylic acids, 10a–g, were generated from the corresponding Na or K salt 9, and obtained in high yields with generally high isotopic enrichment. Although exceedingly high temperatures (280–440 °C) and high pressures (4 atm) were required, this work clearly provided inspiration for more recent developments.

Scheme 2 Thermally induced ¹⁴C isotope exchange of carboxylic salt 9.¹⁹

A few years later, Parnes reported a related strategy for the ¹⁴C labelling of an aryl acetic acid (Scheme 3).²² Though only a single example of isotope exchange was reported, the manuscript included an engaging discussion on mechanisms and concepts for carbon isotope exchange that have been built on in more recent years.

Scheme 3 Synthetic ¹⁴C isotope exchange on aryl acetic acid 11.²²

2.2. Transition metal-promoted carbon isotope exchange

Recent years have seen a significant growth in the development of carbon isotope exchange strategies involving CO₂. This area garnered newfound interest through a series of reports published between 2018 and 2019 by the groups of Audisio, Baran and Martin on transition metal catalysed carbon isotope exchange.^23–25

Audisio and co-workers reported a copper catalysed carbon isotope exchange of aromatic and benzopyrone derived carboxylic acids (Scheme 4).²³ Use of the cesium salt of the acid was necessary to limit competing protodecarboxylation. The conditions and scope of the process reflected related copper promoted decarboxylations of aromatic acids,²⁶ namely that relatively high temperatures (>150 °C) were required and aromatic substrates bearing ortho substituents showed the greatest reactivity. As three equivalents of labelled CO₂ gas was used, the highest possible enrichment at equilibrium is 75%. In general, the isotopic enrichment varied between substrates, with a slight trend that more sterically hindered substrates showed lower isotopic incorporation. The procedure was also applied to the late-stage ¹³C and ¹⁴C isotope labelling of biologically relevant molecules such as probenecid 14g and flumequine 14h. Although mechanistic studies were not conducted, the authors were able to suggest a possible mechanism based on previous knowledge (Scheme 4B).²⁶ Firstly, salt exchange with the copper catalyst (step i) followed by decarboxylation (step ii) provided the aryl copper species II. The labelled CO₂ was then trapped by intermediate II to deliver carboxylate III (step iii). Another salt exchange with the cesium salt 13 delivered the labelled carboxylate 15 (step iv) and regenerated the unlabelled copper carboxylate I to close the catalytic cycle. At the end of the process, the mixture of copper and cesium carboxylates III and 15 were protonated to deliver the isotope enriched carboxylic acids 14 shown in Scheme 4A.

Scheme 4 (A) Cu catalysed ¹³C/¹⁴C isotope exchange on carboxylic salt 13. (B) Proposed mechanism for catalytic cycle of Cu mediated carbon isotope exchange.²³

In 2019, Baran and co-workers demonstrated a nickel mediated carbon isotope exchange of aliphatic carboxylic acids (Scheme 5).²⁴ This process made use of the known reactivity of N-hydroxyphthalimide derived redox active esters 16 as an effective strategy for carboxylic acid activation. The procedure was applicable to a range of primary and secondary carboxylic acids, however, tertiary carboxylic acids were incompatible. A range of medicinally-relevant substrates underwent carbon isotope exchange (e.g. see 17d and 17e) and the process was also applied to ¹³C and ¹⁴C labelling. The isotope enrichment/specific activity was relatively low, however, the authors reasoned that the levels obtained in this study were still of use in early-stage development.

Scheme 5 Ni mediated carbon isotope exchange of aliphatic redox active esters 16.^{24 a} Alternative reaction conditions for ¹⁴C labelling of ¹⁴CO₂ (1 atm), DMF (0.05M), −25 °C (1 h) to room temperature (20 h).

Following on from the work by Baran and co-workers, a related process was reported in the same year by the Martin group (Scheme 6).²⁵ A key development in this study was the ability to use catalytic quantities of the nickel catalyst, rather than the stoichiometric amounts that had previously been required. The isotope incorporation was also generally higher in this report, though consistently high values were unachievable, as is often encountered in carbon isotope exchange processes. To get around this problem, the authors developed an alternative formal CO₂ exchange through a sequence of decarboxylative halogenation and C–X carboxylation.²⁷ This led to complete isotope incorporation in all cases, though the yields were lower. The authors also demonstrated the preparation of the ¹³C and ¹⁴C labelled benzoic acid 19e from the corresponding aryl bromide.²⁸ Finally, the authors revealed the application of this carbon isotope exchange strategy in chain-walking carboxylation reactions (Scheme 6B).²⁹ This process was initiated by conversion of the carboxylic acid 20 to the alkyl bromide 21. Upon exposure to a nickel catalyst, substrates 21 underwent a chain-walking process towards functionalisation at the terminal position of the alkyl chain. Thus, in the presence of ¹³CO₂ the labelled products 22 were formed with good isotopic enrichment, albeit in relatively low yields. Overall, the authors described several strategies for delivering isotope labelled aliphatic and aromatic acids, including medicinally relevant molecules, through formal carbon isotope exchange processes.

Scheme 6 (A) Ni catalysed carbon isotope exchange of carboxylic acid derivatives 18.²⁵ 19e was synthesised from the corresponding aryl bromide using alternative conditions: ^a NiBr₂·dme (10 mol%), neucoproine (20 mol%) in DMA at 50 °C, ^b NiI₂ (2.5 mol%), L3 (4.4 mol%), DMF at 25 °C. (B) Ni catalysed carbon isotope exchange via chain-walking. DBI = dibromoisocyanuric acid.

Neither report by the Baran group or Martin group discussed the mechanism in detail, which is reasonable as the intricacies of these nickel catalysed processes are somewhat disputed.³⁰ In order to inform the reader, we provide simplified mechanisms for these processes, however, we leave our discussion deliberately vague due to the uncertainty surrounding these details (Scheme 7). We suggest that the role of the manganese was to reduce the nickel precatalyst to an active nickel species IV (step i),^31,32 which could promote decarboxylation of the redox active ester 18 to provide the nucleophilic intermediate V (step ii). The labelled CO₂ was then captured by intermediate V to give nickel carboxylate VI (step iii).^33,34 Under Baran's conditions, which used stoichiometric nickel, carboxylate VI would then deliver the desired carboxylic acid 19 upon workup (step iv).²⁴ In the case of Martin and co-workers, the nickel catalyst IV would be regenerated from the reaction between nickel carboxylate VI and manganese metal (step v).²⁵ The manganese carboxylate generated in this step would then lead to the carboxylic acid product 19 upon workup.

Scheme 7 Proposed mechanism for Ni-mediated/catalysed carbon isotope exchange.

2.3. Transition metal-free/thermally induced carbon isotope exchange

In 2020, Audisio and Lundgren concurrently reported the carbon isotope exchange of phenylacetic acids (Scheme 8).^35,36 These methods significantly built on the early work described in section 2.1.^19–22 The procedures involved the mixing of phenylacetate salts 23 in polar aprotic solvent with labelled CO₂. The main difference between the processes was that Audisio and co-workers used cesium carboxylate salts and a small excess of labelled gas (3 equivalents) in DMSO, whereas Lundgren and co-workers used the potassium carboxylate salts with a larger excess of labelled CO₂ in DMF. Notably, Audisio also extended the scope to ¹⁴C and ¹¹C isotope exchange. The conditions reported by Lundgren generally provided higher isotope enrichment, however, this came at the cost of using greater equivalents of labelled CO₂. In both cases, more sterically encumbered substrates tended to give lower isotope enrichment (e.g. see 24a). The required temperatures varied greatly between substrates, for example, 4-cyanophenylacetate 24f readily underwent exchange at 20 °C, whereas 24e required heating at 190 °C. In general, substrates better able to stabilise a build-up of negative charge at the benzylic position proceeded more efficiently. Due to the ubiquity of the phenylacetic acid motif in medicine, both groups were able to demonstrate the reaction with various drug molecules, such as 24d, 24e, 24i and 24j.

Scheme 8 Thermally induced carbon isotope exchange of carboxylate salt 23 from: (A) Audisio and co-workers.³⁵ Standard conditions: *CO₂ (3.0 equiv.), M = Cs, DMSO, 150 °C, 0.5–2 h, then HCl. % incorporation for ¹¹C is based on radiochemical yield: an amount calculated from isotopic purity and corrected for decay. n.r. = not reported. ^a 130 °C. ^b 190 °C. (B) Lundgren and co-workers.³⁶ Standard conditions: *CO₂ (1 atm), M = K, DMF, 160 °C, 48 h, then HCl. ^c 20 °C, 24 h. ^d 40 °C, 24 h. ^e 110 °C, 14 h. (C) Mechanistic proposal for thermally induced carbon isotope exchange.

In terms of the mechanism (Scheme 8C), both groups proposed that decarboxylation could provide the benzylic intermediate VII (step i), which could react directly with labelled CO₂ to provide the desired product 24 (step ii). Alternatively, and building on suggestions put forward by Parnes,²² the benzyl anion VII could instead deprotonate the carboxylate 23 to give the dienolate VIII (step iii). Trapping of this intermediate with the labelled CO₂ would provide the malonate derivative IX (step iv). In the presence of a proton source, either during work up or from another molecule of carboxylate 23, intermediate IX would then undergo decarboxylation (step v) to provide the desired product 24 after protonation.

Building on this work, Lundgren, Rotstein and co-workers then reported a carbon isotope exchange of amino acids (Scheme 9).³⁷ The process involved heating amino acids 25 in the presence of a catalytic amount of 4-anisaldehyde, Cs₂CO₃ and labelled CO₂ gas. A range of Boc-protected labelled proteinogenic and non-proteinogenic amino acids 26 were isolated in generally good to excellent yields. As with most dynamic exchange processes, the isotope incorporation was variable, but the authors reasoned that the levels of enrichment were still sufficient for applications, such as ADME studies. Furthermore, the authors noted that the products can be resubmitted to the exchange conditions to further enhance the isotope enrichment. The catalytic conditions were applicable to exchange with ¹³CO₂ and ¹⁴CO₂. Labelling with ¹¹CO₂ was also possible, however, considering the short half-life of this molecule (t_1/2 = 20 mins), a stoichiometric amount of benzaldehyde was used to deliver sufficient isotope incorporation within 10 minutes.

Scheme 9 (A) Carbon isotope exchange of amino acids 25.^{37 a} (1) PhCHO (1.0 equiv.), Cs₂CO₃ (0.50 equiv.), MeOH, 70 °C, (2) ¹¹CO₂ (7–10 GBq), DMSO, 90 °C, 10 min, then HCl. % incorporation for ¹¹C is based on radiochemical yield: an amount calculated from isotopic purity and corrected for decay. n.r. = not reported. (B) Mechanistic proposal for carbon isotope exchange of amino acids.

The authors again considered a direct route to isotope exchange via an aza-allyl anion (cf. VII, Scheme 8C), but mechanistic studies suggested the alternative route via a malonate derivative (XIII, Scheme 9B; cf. IX, Scheme 8C) was in operation. Thus, the authors proposed that the amino acid 25 first condenses with the corresponding aldehyde to form the imine XI (step i). Tautomerisation (or deprotonation) and carboxylation then delivered the imino malonate derivative XIII (steps ii and iii). Finally, subsequent decarboxylation (step iv) and hydrolysis (step v) delivered the desired labelled amino acid 26 and regenerated the aldehyde catalyst (Scheme 9B).

During their studies on the carbon isotope exchange of amino acids (Scheme 9), Lundgren, Rotstein and co-workers noted the drawback that, unless subsequent resolution was performed, access to enantiopure labelled amino acids was not possible.³⁷ To address this limitation, Lundgren, Derdau and colleagues presented a related carbon isotope exchange for accessing enantioenriched labelled amino acids (Scheme 10).³⁸ The overall strategy involved combining their work on aldehyde-promoted carbon isotope exchange with a chiral aldehyde-promoted stereoconversion of amino acids previously reported by Chin, Kim and co-workers.³⁹ The chiral aldehyde (S)-1 was found most suitable for promoting isotope exchange and controlling the stereochemistry of the product. The authors attempted to use catalytic amounts of chiral aldehyde (S)-1, but ultimately these attempts were fruitless. The reaction was applicable to a variety of natural and non-natural amino acids and both ¹³C and ¹⁴C labelling was achieved, though the scope seemed less general in comparison to their previous racemic approach (cf. Scheme 9). As has become standard in CO₂ exchange processes, the level of isotope incorporation varied between substrates. Due to the lack of methods for generating chiral labelled amino acids, this chemistry demonstrated a particularly elegant solution. However, the enantiomeric ratios of the products were often not ideal (generally <90 : 10 er), thus the authors reverted to resolution techniques for delivering products with high enantiomeric ratios.

Scheme 10 (A) Chiral aldehyde-promoted carbon isotope exchange of amino acids. (B) Proposed diastereoselective mechanism for formation of 28.³⁸

The authors proposed that the CO₂ exchange and stereochemical induction processes proceed in separate steps. Firstly, condensation with the chiral aldehyde (S)-1 provided an imine intermediate that could undergo carbon isotope exchange in a similar fashion to that described in Scheme 9B. This would provide the labelled imine intermediate XV (Scheme 10B, cf. XIV Scheme 9B). In agreement with the reports by Chin, Kim and co-workers,³⁹ the authors suggested that deprotonation of chiral imine XV provided intermediate XVI (step i), which is then susceptible to a diastereoselective reprotonation to give the enantioenriched compound XVII (step ii). Subsequent hydrolysis then delivered the enantioenriched labelled amino acid 28 (step iii).

2.4. Light mediated carbon isotope exchange

In 2021, the groups of Audisio, and Lundgren and Rotstein reported related photoredox catalysed carbon isotope exchange processes in unison.^40,41 The general conditions for both processes were a photoredox catalyst (either 4-CzIPN (PC1) or 4-CzBnBN (PC2)), a base (K₃PO₄ or Cs₂CO₃) and a polar aprotic solvent (DMF, DMA or DMSO) under blue light irradiation (Scheme 11). Both groups realised that 4-CzIPN was converted to 4-CzBnBN under the reaction conditions, in agreement with previous reports.⁴² Thus, whereas 4-CzIPN (PC1) was used in many cases, 4-CzBnBN (PC2) provided a more reliable and efficient exchange process for more complex substrates, as it was less prone to side reactions. This was of note for Lundgren and Rotstein, who discovered that 4-CzBnBN (PC2) could facilitate ¹¹C transfer when 4-CzIPN (PC1) failed. A clear advantage of these processes in comparison to the thermally induced methods (up to 190 °C, see Scheme 8) is the ability to perform exchange at room temperature. Both groups demonstrated reactivity with various phenylacetic acids with known bioactivity (30f, 30l, 30m) and Audisio and coworkers also applied the process in exchange with ¹⁴CO₂ (30f).⁴⁰ Realising that this photoredox catalysed process proceeded much faster than the thermally induced version, Lundgren, Rotstein and co-workers nicely demonstrated the suitability of the reaction in ¹¹C isotope exchange (30l, 30m).⁴¹

Scheme 11 Photocatalyzed carbon isotope exchange of carboxylic acid 29 from: (A) Audisio and co-workers.⁴⁰ Standard conditions: *CO₂ (3.0 equiv.), PC1 (6 mol%), K₃PO₄ (1.0 equiv.), DMF, 6 h, then HCl. ^a PC2 used instead of PC1. (B) Lundgren, Rotstein and co-workers.⁴¹ Standard conditions: *CO₂ (7.0 equiv.), PC1 or PC2 (5 mol%), Cs₂CO₃ (1.0 equiv.), DMA or DMSO, rt. % incorporation for ¹¹C is based on Radiochemical yield: an amount calculated from isotopic purity and corrected for decay. n.r. = not reported. (C) Proposed mechanism for photocatalyzed carbon isotope exchange of carboxylic acid 29.

Both groups provided similar mechanistic proposals, with the Audisio group conducting some informative mechanistic studies to support their suggestions (Scheme 11C).⁴⁰ Previous studies on the photoredox catalysed decarboxylation of phenylacetic acids were also used as support for the proposed mechanisms.⁴² Under basic reaction conditions, the carboxylic acid 29 was initially deprotonated to give carboxylate XVII (step i). Oxidation of the carboxylate by the photoredox catalyst then provided the carboxyl radical XVIII (step ii), which was prone to decarboxylation towards carbon centred radical XIX (step iii). Single electron transfer from the reduced photoredox catalyst to the radical species XIX would then provide the nucleophile XX (step iv), which was able to capture the labelled CO₂ gas towards the desired product 30 (step v). In contrast to the mechanistic proposals for thermally induced carbon isotope exchange in which a malonate intermediate was proposed (cf. intermediate IX in Scheme 8 and XIII in Scheme 9), the direct capture of labelled CO₂ gas by the benzyl anion XX (step v) was suggested in this study.

Following these reports, Kong and co-workers expanded the methodology towards a formal carbon isotope exchange of α-keto acids via thioketal acid salts 31 (Scheme 12).⁴³ The reaction was applied to a range of substrates, with the yields and levels of isotope enrichment consistent with the efficiencies observed in related photoredox catalysed isotope exchanges (cf. Scheme 11).^40,41 The preparation of isotope-labelled drug derivatives (32i, 32j) was also demonstrated. Mechanistic studies, including Stern–Volmer quenching experiments, supported a mechanistic pathway like that shown in Scheme 11C.

Scheme 12 Photocatalysed carbon isotope exchange of protected α-keto acids. ^a Isolated as the corresponding methyl ester.⁴³

The Kong group were then able to demonstrate a direct carbon isotope exchange of α-ketoacids, foregoing the steps of thioketal installation and removal, that were previously required (Scheme 13).⁴⁴ The reactivity was once again demonstrated on a range of substrates, though overall the scope, yields and isotope enrichment were diminished in comparison to their previous report. Nonetheless, by developing both an indirect but general (Scheme 12) and a more direct yet less efficient (Scheme 13) CO₂ exchange, a flexible approach to labelled α-ketoacids was demonstrated. Interestingly, this photoinitiated process did not require a photoredox catalyst. Through several preliminary experiments, the authors tentatively proposed the mechanism described in Scheme 13B. They suggested irradiation with light forms the excited state of the α-keto carboxylate XXII (step i), which was able to undergo single electron transfer with the ground state of another molecule of α-ketoacid 33 (step ii). Decarboxylation of intermediate XXIV would generate the radical species XXV (step iii), which could then undergo single electron transfer with XXIII to provide the acyl anion XXVI (step iv) and regenerate the ground state α-keto acid XXI to maintain the catalytic cycle. Carboxylation of the acyl anion XXVI then would deliver the labelled a-keto ester 34 (step v). Further studies would better delineate the mechanism of this interesting process.

Scheme 13 (A) Direct carbon isotope exchange of α-keto acids. ^a CO₂ (3.0 equiv.), CuCN/BINAP (20 mol%). (B) Mechanistic proposal.⁴⁴

3. CO₂ transfer reactions

3.1. Early reports on CO₂ transfer reactions

The earliest documentation of a CO₂ transfer reaction was outlined by Raecke in 1958 (Scheme 14).⁴⁵ The process involved a disproportionation reaction of potassium benzoate 35 into terephthalate 36 and benzene 37 (Scheme 14). This type of disproportionation reaction is sometimes referred to as the Henkel or Raecke reaction. The method required heating to very high temperatures (350–500 °C), often in the presence of a cadmium catalyst. The harsh conditions limit the generality of the process, but it has found interest in the preparation of bulk chemicals, such as terephthalic acid and furan-2,5-dicarboxylic acid.^46,47 The mechanism has not been well studied, but it is proposed that heating promotes decarboxylation of the benzoic acid salt 35 to give the anionic species XXVII (step i). The intermediate XXVII then deprotonates another molecule of the aromatic acid salt 35. This produces benzene 37 and the intermediate XXVIII which is able to capture the CO₂ that was expelled during the first decarboxylation step. Though an impressive transformation, the harsh conditions have prevented the widespread use of this chemistry. In addition, many of the processes also required the addition of (pressurised) CO₂, thus a true CO₂ transfer process is not in operation.⁴⁸

Scheme 14 The Henkel process.⁴⁵

3.2. CO₂ transfer via addition across alkenes

An impressive example of CO₂ transfer was reported by Yu and co-workers in 2021 (Scheme 15A).⁴⁹ The process involved the addition of the C–CO₂H bond in amino acids across the C=C double bond of activated alkenes. Importantly, the work demonstrated the efficient transfer of CO₂ under mild reaction conditions. The reaction was generally performed using cyclic amino acids and styrene derivatives (see 40a, 40b and 40d), though extending the reactivity to natural amino acid derivatives and other alkenes (see 40e) was also possible. The reaction with several drug derivatives (see 40c) was also demonstrated. The labelled product 40f was prepared using the corresponding labelled amino acid. Though this was used as a mechanistic experiment, it does offer a possible route to apply this method in isotope labelling.

Scheme 15 (A) CO₂ transfer from amino acids across alkenes. ^aCsF (3.0 equiv.).⁴⁹ (B) Mechanistic proposal.

Several other mechanistic studies were conducted to support the proposed mechanism of this photoredox catalysed process (Scheme 15B), which the authors suggested to involve an initial oxidation of the carboxylate XXIX by the excited state of the photoredox catalyst (step ii). This provided a carboxyl radical which was prone to decarboxylation, leading to the alkyl radical XXX. Addition to the alkene 38 then afforded the radical species XXXI (step iii), which could undergo single electron transfer with the reduced state of the photoredox catalyst to give the anionic species XXXII and close the catalytic cycle (step iv). Finally, the intermediate XXXII would capture the in situ generated CO₂ to provide the desired product 40 after protonation.

3.3. CO₂ transfer via deprotonation

A general strategy for CO₂ transfer was reported by Perry, Yorimitsu and co-workers who developed the salt of triphenylacetic acid 42-K for the transfer of CO₂ to a range of compounds, including (hetero)arenes, alkynes and amines (Scheme 16A).⁵⁰ By using the labelled analogue of reagent 42-K, the process was also extended to the isotope labelling of some biologically relevant molecules (see 43k, 43l). Importantly, all labelled products were isolated with high isotope enrichment, a general advantage of transfer reactions over carbon isotope exchange. Interestingly, reagent 42-K provided the source of base and CO₂ for the reaction, leading the authors to describe these compounds as dual-function reagents.

Scheme 16 (A) CO₂ transfer via decarboxylative metalation.⁵⁰ (B) Mechanistic proposal.

With the support of mechanistic studies (Scheme 16B), they proposed that reagent 42-K initially underwent decarboxylation to provide CO₂ and the relatively strong base XXXIII (step i, pK_a of Ph₃CH = 30.6 in DMSO).⁵¹ This enabled the deprotonation of the substrate 41 to provide the intermediate XXXIV (step ii), which was then able to capture the in situ generated CO₂ and deliver the carboxylated product 43 after alkylation (step iii). In general, substrates containing a C–H bond with a pK_a of 30.6 or less (in DMSO) were able to undergo an efficient carboxylation, reflecting the basicity of the intermediate XXXIII.

Building on the work by Perry and Yorimitsu (Scheme 16)⁵⁰ and their earlier reports on the formation of labelled α-ketoacids (Scheme 12),⁴³ Kong and co-workers revealed the application of this reactivity in a CO₂ transfer with benzaldehyde-derived dithianes 44 (Scheme 17).⁵² A variety of functional groups were compatible (45a–h) and the reactivity was extended to the carboxylation of selected toluene derivatives (45i, 45j). Various drug derivatives (45k, 45l) also performed well in this CO₂ transfer and the extension of this method to isotope labelling (45a, 45b and 45k) was demonstrated. As generally seen in CO₂ transfer reactions, the level of isotope incorporation was high in all cases. The mechanism for the process likely proceeds in a similar fashion to that shown in Scheme 16B.

Scheme 17 CO₂ transfer with benzaldehyde-derived dithianes and toluene derivatives.⁵²

Shigeno and co-workers have developed an alternative silicon-based reagent 47 for CO₂ transfer reactions (Scheme 18).⁵³ This work built on their/Kondo's previous work in carboxylation chemistry⁵⁴ by applying it to the arena of CO₂ transfer. The reaction was demonstrated on the C–H carboxylation of (hetero)arenes (48a–i) and in benzylic carboxylation (48j). As the base (Et₃COCs, 50; pK_a of t-BuOH = 32.2 in DMSO)⁵¹ generated during this process is more reactive than that formed in the method reported by Yorimitsu and Perry (Ph₃CK, 42-K; pK_a of Ph₃CH = 30.6 in DMSO; Scheme 16)⁵¹ it has the potential to offer wider substrate scope, however, at present, the scope of each process is relatively similar. By using the isotope labelled reagent 47, the procedure was also used to prepare the isotope labelled compound 48i. In addition to the carboxylating agent 47, CsF and 18-crown-6 were also required to promote the reaction. The reaction mechanism was suggested to proceed in a similar fashion to that shown in Scheme 16, except that the CO₂ and active base (Et₃COCs, 50) were proposed to form upon desiliylation of the reagent 47 with CsF via intermediate 49 (Scheme 18B).

Scheme 18 CO₂ transfer reaction utilizing silicon-based reagents.⁵³

To further illustrate the power of dual-function reagents in carboxylation reactions, Perry, Kong and co-workers have recently described a Kolbe–Schmitt-type carboxylation of phenols (Scheme 19).⁵⁵ The procedure overcame some of the longstanding challenges associated with Kolbe–Schmitt type reactivity, such as excessive temperatures and pressures. Interestingly, by using the cesium salt of the carboxylating agent 42-Cs (and in some cases the potassium salt 42-K) unique para-selectivity was observed. The sodium salt of the carboxylating agent 42-Na, provided ortho-carboxylation in line with classical Kolbe–Schmitt-type reactivity, however, the yield of this process was low (38%). Access to a range of substituted 4-hydroxybenzoic acids was demonstrated with high para-selectivity (52a–52g). ortho-Carboxylation was observed in some cases, but only when the para-position was blocked or sterically encumbered (see 52h). Owing to the ability to use near-stoichiometric quantities of the carboxylating reagent, this approach offered a practical and efficient route to ¹³C-labelled 4-hydroxybenzoic acid derivatives (52a, 52c).

Scheme 19 para-Selective Kolbe–Schmitt reaction.⁵⁵

Recently, the Perry and Kong groups concurrently reported the application of CO₂ transfer reactions in regiodivergent indole functionalization (Scheme 20).^56,57 Both groups found that N–H carboxylation occurred at lower temperatures (40–50 °C) whereas C3–H carboxylation was possible at higher temperatures (120–140 °C). In addition, the equivalents of the carboxylating agent 42-K or 42-Cs were increased when switching between N–H (1.1 equivalents) and C3–H carboxylation (2.0–2.2 equivalents). The Perry group used the potassium salt of triphenylacetic acid (Ph₃CCO₂K, 42-K, Scheme 20A), whereas the Dong group opted for the cesium salt (Ph₃CCO₂Cs, 42-Cs, Scheme 20B), but overall, there was no significant difference in reactivity between the two salts. Both groups presented a detailed scope representing switchable N–H vs. C3–H carboxylation on a range of indoles, 54/55a–g, with applications in late-stage ¹³C labelling and the carboxylation of biologically relevant compounds, 55a. Perry and co-workers also demonstrated a series of related transformations with indoles and the N–H carboxylation of other amines. Both groups offered similar mechanistic hypotheses following the types of steps described in Scheme 16B. Here (Scheme 20C), upon generation of the trityl anion XXXIII, the indole is deprotonated to provide the intermediate XXXV. The N–H carboxylated product 54 can then be accessed when conducting the experiment at low temperatures and lower equivalents of the carboxylating agent 42-K/42-Cs. Alternatively, by increasing the reaction temperature and the number of equivalents of the carboxylating agent 42-K/42-Cs the C3–H carboxylated product 55 is produced.

Scheme 20 Switchable N–H and C3–H CO₂ transfer on indoles by: (A) Perry and co-workers.^{56 a} Standard conditions: 42-K (1.1 equiv.), DMF (0.2 M), 50 °C, 4 h, then alkylation. ^b 42-K (2.0 equiv.), 140 °C. (B) Kong and co-workers.^{57 c} Standard conditions: 42-Cs (1.1 equiv.), DMF (0.2 M), 40 °C, 12 h, then alkylation. ^d 42-Cs (2.2 equiv.), 120 °C (C) Proposed mechanism for switchable reactivity.

3.4. CO₂ transfer via decarboxylative metal–halogen exchange

Until this point, the CO₂ transfer reagents described in Schemes 16–20 promoted carboxylation through a sequence of deprotonation and carboxylation. To further illustrate the abilities of these reagents, Perry, Yamazaki and co-workers developed a C–X carboxylation of (hetero)aromatic halides (Scheme 21A).⁵⁸ The reaction was applicable to either heteroaryl bromides or iodides, however, chlorides were unreactive under the reaction conditions. Through a series of experimental and computational studies, the authors suggested a mechanism similar to that shown in Scheme 16B in which the dual-function reagent 42-K acts as the source of CO₂ and metalating agent. A key advancement in this work was that the metalated intermediate XXXIV was generated through metal-halogen exchange, rather than deprotonation (cf. Scheme 16B, step ii). Noteworthy features included: (1) the site of carboxylation was governed by the position of the C–X bond, for example, 2- and 5-brominated thiazoles 56d and 56e gave the 2- and 5-carboxylated products 57d and 57e. (2) Through examination of the reaction scope and a robustness screen,⁵⁹ the reaction was shown to be tolerant of various functional groups that are usually incompatible with traditional metal–halogen exchange processes, such as aldehydes (57f), ketones, amides, alkyl chlorides and nitriles. (3) The process was applied in the isotope labelling of some biologically relevant molecules, for example the Edoxaban precursor 57g. Finally, the authors demonstrated a formal carbon isotope exchange (Scheme 21B). In this process the drug molecule Febuxostat 57h was first converted into the aryl bromide 57h via decarboxylative bromination,⁶⁰ followed by C–X carboxylation under the standard reaction conditions.

Scheme 21 (A) CO₂ transfer via decarboxylative metal–halogen exchange.⁵⁸ (B) Formal late-stage carbon isotope exchange.

4. Conclusions

Recent years have seen the rise of carbon isotope exchange and CO₂ transfer reactions as effective strategies in the field of carboxylation. These methods advance the current-state-of-the-art, in particular towards the development of efficient carbon isotope labelling processes. We have provided a thorough review covering early studies to the latest developments in these areas and highlighted the advantages and limitations of each strategy. Developments in both areas have uncovered novel reactivity that has advanced the field of carboxylation. Further development of these processes is highly anticipated, owing to the key role carbon isotope labelling has in the pharmaceutical and agrochemical fields. Expanding the generality in both carbon isotope exchange and transfer reactions with CO₂ is required to make these processes more general and reliable. In particular, further application of CO₂ transfer reactions in late-stage-functionalization of complex molecules is anticipated. The robustness of carbon isotope exchange reactions could also be improved as many examples often give low chemical yields, even if % isotope incorporation is satisfactory. Both processes could be made more amenable to ¹¹C isotope labelling. The application of CO₂ transfer in ¹¹C isotope labelling seems distant at present due to the need to prepare these reagents and the short lifetime of the ¹¹C isotope, however, advancements in automation may aid this area.¹⁸ Although some examples of ¹¹CO₂ incorporation via carbon isotope exchange are known they often result in low molar activity which limits possible applications. The inherent nature of the process (a direct competition for carboxylation with either labelled or non-labelled CO₂, see Scheme 1B) is somewhat to blame for this, however, solutions to this problem, such as faster reactions and greater levels of exchange, would greatly improve applicability. Overall, we hope this review informs the community of the benefits and limitations of carbon isotope exchange and CO₂ transfer reactions and encourages others to consider these strategies when planning their own carboxylation or carbon isotope labelling reactions.

Conflicts of interest

There are no conflicts to declare.

Data availability

No primary research results, software or code have been included and no new data were generated or analysed as part of this review.

Acknowledgements

This work was supported by a UKRI New Investigator Award (UKRI1101, APP24094) and an EPSRC ICASE award in collaboration with AstraZeneca (WT8946489).

References

1. R. Voges, J. R. Hayes and T. Moenius, in Preparation of Compounds Labeled with Tritium and Carbon-14, Wiley, Chichester, UK, 2009.
2. C. S. Elmore and R. A. Bragg, Isotope chemistry; a useful tool in the drug discovery arsenal, Bioorg. Med. Chem. Lett., 2015, 25, 167–171.
3. (a) J. Andersson, P. Truong and C. Halldin, In-target produced [¹¹C]methane: Increased specific radioactivity, Appl. Radiat. Isot., 2009, 67, 106–110; (b) Z. Debreczeni-Máté, O. Freihat, I. Törő, M. Simon, Á. Kovács and D. Sipos, Value of ¹¹C-Methionine PET Imaging in High-Grade Gliomas: A Narrative Review, Cancers, 2024, 16, 3200.
4. T. H. Witney and K. M. Brindle, Imaging tumour cell metabolism using hyperpolarized ¹³C magnetic resonance spectroscopy, Biochem. Soc. Trans., 2010, 38, 1220–1224.
5. V. Derdau, C. S. Elmore, T. Hartung, B. McKillican, T. Mejuch, C. Rosenbaum and C. Wiebe, The Future of (Radio)-Labeled Compounds in Research and Development within the Life Science Industry, Angew. Chem., Int. Ed., 2023, 62, e202306019.
6. E. M. Isin, C. S. Elmore, G. N. Nilsson, R. A. Thompson and L. Weidolf, Use of Radiolabeled Compounds in Drug Metabolism and Pharmacokinetic Studies, Chem. Res. Toxicol., 2012, 25, 532–542.
7. S. Gehen, M. Corvaro, J. Jones, M. Ma and Q. Yang, Challenges and Opportunities in the Global Regulation of Crop Protection Products, Org. Process Res. Dev., 2019, 23, 2225–2233.
8. L. Faba and S. Ordóñez, Carboxylation reactions for the sustainable manufacture of chemicals and monomers, RSC Sustain., 2024, 2, 3167–3182.
9. V. Babin, F. Taran and D. Audisio, Late-Stage Carbon-14 Labeling and Isotope Exchange: Emerging Opportunities and Future Challenges, JACS Au, 2022, 2, 1234–1251.
10. The mechanisms depicted in Scheme 1B and C are simplified. We have shown intermediates 3 and 8 as anionic species for simplicity. Radical and organometallic intermediates are also possible intermediates in these reactions.
11. A handful of reports on carbon isotope exchange comment on the capture of excess CO₂. However, quantifying the amount of recovered CO₂ has not been described. We quote directly from the original reports: ref. 14 – “The ¹⁴CO₂ pressure was about 4–5 atm at the reaction temp. The system was opened and ¹⁴CO₂ recovered in the form of Ba¹⁴CO₃. Ref. 19 – In the SI: “Question 14: Can unreacted ¹⁴CO₂ in the reaction flask be recaptured for use in other reactions? Answer: Efforts to capture ¹⁴CO₂ from the side-arm flask into base solution had limited success. No attempt to capture the ¹⁴CO₂ by recondensation with liquid nitrogen was attempted. HPLC analyses of crude reaction mixtures with radioactivity detection does suggest that a significant portion of the unreacted ¹⁴CO₂ remains bound to the nickel catalyst. Ref. 32 – In the SI: “After stirring for 17 hours, the excess [¹³C]CO₂ was transferred into a separate flask (flask 3) filled with an aqueous 2% BaOH solution (15 mL), while being cooled with liquid nitrogen”.
12. For reviews on carbon isotope exhange, see: (a) A. Labiche, A. Malandain, M. Molins, F. Taran and D. Audisio, Modern Strategies for Carbon Isotope Exchange, Angew. Chem., Int. Ed., 2023, 62, e202303535; (b) K. Hinsinger and G. Pieters, The Emergence of Carbon Isotope Exchange, Angew. Chem., Int. Ed., 2019, 58, 9678–9680; (c) C. S. Elsmore, Use of Isotopically Labeled Compounds in Drug Discovery, Annu. Rep. Med. Chem., 2009, 44, 515–534; (d) R. A. Bragg, M. Sardana, M. Artelsmair and C. S. Elmore, New trends and applications in carboxylation for isotope chemistry, J. Labelled Compd. Radiopharm., 2018, 61, 934–948.
13. For reviews on broader applications of carbon isotope chemistry, see: (a) D. U. Nielsen, K. T. Neumann, A. T. Lindhardt and T. Skrydstrup, Recent developments in carbonylation chemistry using [¹³C]CO, [¹¹C]CO, and [¹⁴C]CO, J. Labelled Compd. Radiopharm., 2018, 61, 949–987; (b) V. Derdau, New trends and applications in cyanation isotope chemistry, J. Labelled Compd. Radiopharm., 2018, 61, 1012–1023; (c) T. J. Gregson, J. M. Herbert and E. C. Row, Synthetic approaches to regiospecifically mono- and dilabelled arenes, J. Labelled Compd. Radiopharm., 2011, 54, 1–32; (d) A. Del Vecchio, G. Destro, F. Taran and D. Audisio, Recent developments in heterocycle labeling with carbon isotopes, J. Labelled Compd. Radiopharm., 2018, 61, 988–1007.
14. A. Pees, M. Chassé, A. Lindberg and N. Vasdev, Recent Developments in Carbon-11 Chemistry and Applications for First-In-Human PET Studies, Molecules, 2023, 28, 931.
15. During the preparation of this manuscript, Kong and co-workers published a related review on CO₂ transfer/shuttling; X. Liu, H. Wang and D. Kong, CO₂ shuttling in organic synthesis, Chem. Synth., 2025 DOI:10.20517/cs.2025.51.
16. For reviews on isodesmic reactions and shuttle catalysis, see: (a) B. N. Bhawal and B. Morandi, Isodesmic Reactions in Catalysis – Only the Beginning?, Isr. J. Chem., 2018, 58, 94–103; (b) B. N. Bhawal and B. Morandi, Shuttle Catalysis—New Strategies in Organic Synthesis, Chem. – Eur. J., 2017, 23, 12004–12013; (c) B. N. Bhawal and B. Morandi, Catalytic Isofunctional Reactions—Expanding the Repertoire of Shuttle and Metathesis Reactions, Angew. Chem., Int. Ed., 2019, 58, 10074–10103.
17. For reviews on carbon-based functional group transfer reactions see: (a) R. Chinchilla and A. Baeza, Hydrocyanation of Non-Activated Alkenes Without Handling HCN, Eur. J. Org. Chem., 2024, e202400527; (b) B. N. Bhawal and B. Morandi, Catalytic Transfer Functionalization through Shuttle Catalysis, ACS Catal., 2016, 6, 7528–7535; (c) J. Cao, Z.-J. Zheng, Z. Xu and L.-W. Xu, Transition-metal-catalyzed transfer carbonylation with HCOOH or HCHO as non-gaseous C1 source, Coord. Chem. Rev., 2017, 336, 43–53; (d) L. Wu, Q. Liu, R. Jackstell and M. Beller, Carbonylations of Alkenes with CO Surrogates, Angew. Chem., Int. Ed., 2014, 53, 6310–6320.
18. M. Molins, A. Hauwelle, L. Fogel, Q. Lemesre, O. Loreau, F. Taran, F. Caillé and D. Audisio, Electrophilic Cyanation with Carbon Isotopes: A Versatile Access to ¹³C, ¹⁴C, and ¹¹C Aryl Nitriles from Labeled NCTS, J. Am. Chem. Soc., 2025, 147, 35031–35041.
19. A. Szabolcs, J. Szammer and L. Noszkó, A new method for the preparation of carboxyl-labelled aliphatic carboxylic acids, Tetrahedron, 1974, 30, 3647–3648.
20. R. Nakai, M. Sugii and H. Nakao, Isotopic Tracer Studies of the Ketonic Pyrolysis of Sodium Carboxylates, J. Am. Chem. Soc., 1959, 81, 1003–1006.
21. D. J. Darensbourg, E. M. Longridge, M. W. Holtcamp, K. K. Klausmeyer and J. H. Reibenspies, Reversible decarboxylation of phosphine derivatives of Cu(I) cyanoacetate. Mechanistic aspects germane to catalytic decarboxylation of carboxylic acids, J. Am. Chem. Soc., 1993, 115, 8839–8840.
22. H. Parnes, A method for the preparation of 14C-labeled carboxylic acids. Synthesis of 6,11-dihydro[b,e]thiepin-11-one-3-yl acetic ¹⁴C-acid, J. Labelled Compd. Radiopharm., 1979, 16, 771–775.
23. G. Destro, O. Loreau, E. Marcon, F. Taran, T. Cantat and D. Audisio, Dynamic Carbon Isotope Exchange of Pharmaceuticals with Labeled CO₂, J. Am. Chem. Soc., 2019, 141, 780–784.
24. C. Kingston, M. A. Wallace, A. J. Allentoff, J. N. deGruyter, J. S. Chen, S. X. Gong, S. Bonacorsi, Jr. and P. S. Baran, Direct Carbon Isotope Exchange through Decarboxylative Carboxylation, J. Am. Chem. Soc., 2019, 141, 774–779.
25. A. Tortajada, Y. Duan, B. Sahoo, F. Cong, G. Toupalas, A. Sallustrau, O. Loreau, D. Audisio and R. Martin, Catalytic Decarboxylation/Carboxylation Platform for Accessing Isotopically Labeled Carboxylic Acids, ACS Catal., 2019, 9, 5897–5901.
26. For transition-metal promoted decarboxylations of aromatic acids, see: (a) L. J. Gooßen, W. R. Thiel, N. Rodríguez, C. Linder and B. Melzer, Copper-Catalyzed Protodecarboxylation of Aromatic Carboxylic Acids, Adv. Synth. Catal., 2007, 349, 2241–2246; (b) L. J. Gooßen, G. Deng and L. M. Levy, Synthesis of Biaryls via Catalytic Decarboxylative Coupling, Science, 2006, 313, 662–664; (c) G. J. P. Perry and I. Larrosa, Recent Progress in Decarboxylative Oxidative Cross-Coupling for Biaryl Synthesis, Eur. J. Org. Chem., 2017, 3517–3527; (d) M. Font, J. M. Quibell, G. J. P. Perry and I. Larrosa, The use of carboxylic acids as traceless directing groups for regioselective C–H bond functionalisation, Chem. Commun., 2017, 53, 5584–5597.
27. For related formal isotope exchange procedures, see: (a) C. Brown, J. Eustache, J. P. Frideling and B. Shroot, Synthesis of ¹⁴C-anthralin, J. Labelled Compd. Radiopharm., 1984, 21, 973–983; (b) D. R. Hicks, Synthesis of [¹⁴C]-labelled tolrestat(N-[[5-(trifluoromethyl)-6-methoxy-1-naphthalenyl]-[¹⁴C]thioxomethyl]-N-methylglycine; AY-27, 773), J. Labelled Compd. Radiopharm., 1984, 21, 229–235; (c) S. Ravi, K. M. Mathew, V. K. P. Unny and N. Sivaprasad, A facile synthesis of high specific activity sodium [1-14C] lauryl sulphate under microwave irradiation, J. Labelled Compd. Radiopharm., 2005, 48, 1055–1058.
28. The authors used the aryl bromide in this case and a formal isotope exchange was not demonstrated. However, an isotope exchange could be envisioned by conducting a decarboxylative halogenation as the initial step, e.g. see ref. 60.
29. (a) F. Juliá-Hernández, T. Moragas, J. Cornella and R. Martin, Remote carboxylation of halogenated aliphatic hydrocarbons with carbon dioxide, Nature, 2017, 545, 84–88; (b) H. Sommer, F. Juliá-Hernández, R. Martin and I. Marek, Walking Metals for Remote Functionalization, ACS Cent. Sci., 2018, 4, 153–165; (c) C. Romano and R. Martin, Ni-catalysed remote C(sp³)-H functionalization using chain-walking strategies, Nat. Rev. Chem., 2024, 8, 833–850.
30. (a) J. B. Diccianni and T. Diao, Mechanisms of Nickel-Catalyzed Cross-Coupling Reactions, Trends Chem., 2019, 1, 830–844; (b) A. Bismuto, P. Finkelstein, P. Müller and B. Morandi, The Journey of Ni(I) Chemistry, Helv. Chim. Acta, 2021, 104, e2100177; (c) G. A. Dawson, E. H. Spielvogel and T. Diao, Nickel-Catalyzed Radical Mechanisms: Informing Cross-Coupling for Synthesizing Non-Canonical Biomolecules, Acc. Chem. Res., 2023, 56, 3640–3653; (d) D. A. Cagan, D. Bím, N. P. Kazmierczak and R. G. Hadt, Mechanisms of Photoredox Catalysis Featuring Nickel–Bipyridine Complexes, ACS Catal., 2024, 14, 9055–9076.
31. N. Hazari, P. R. Melvin and M. M. Beromi, Well-defined nickel and palladium precatalysts for cross-coupling, Nat. Rev. Chem., 2017, 1, 0025.
32. A. Correa, T. León and R. Martin, Ni-Catalyzed Carboxylation of C(sp²)– and C(sp³)–O Bonds with CO₂, J. Am. Chem. Soc., 2014, 136, 1062–1069.
33. R. J. Somerville, C. Odena, M. F. Obst, N. Hazari, K. H. Hopmann and R. Martin, Ni(I)–Alkyl Complexes Bearing Phenanthroline Ligands: Experimental Evidence for CO₂ Insertion at Ni(I) Centers, J. Am. Chem. Soc., 2020, 142, 10936–10941.
34. R. J. Somerville and R. Martin, in Nickel Catalysis in Organic Synthesis, Wiley-VCH, 2020, pp. 285–330.
35. G. Destro, K. Horkka, O. Loreau, D.-A. Buisson, L. Kingston, A. Del Vecchio, M. Schou, C. S. Elmore, F. Taran, T. Cantat and D. Audisio, Transition-Metal-Free Carbon Isotope Exchange of Phenyl Acetic Acids, Angew. Chem., Int. Ed., 2020, 59, 13490–13495.
36. D. Kong, P. J. Moon, E. K. J. Lui, O. Bsharat and R. J. Lundgren, Direct reversible decarboxylation from stable organic acids in dimethylformamide solution, Science, 2020, 369, 557–561.
37. O. Bsharat, M. G. J. Doyle, M. Munch, B. A. Mair, C. J. C. Cooze, V. Derdau, A. Bauer, D. Kong, B. H. Rotstein and R. J. Lundgren, Aldehyde-catalysed carboxylate exchange in α-amino acids with isotopically labelled CO₂, Nat. Chem., 2022, 14, 1367–1374.
38. M. G. J. Doyle, O. Bsharat, A. Sib, V. Derdau and R. J. Lundgren, Enantioselective Carbon Isotope Exchange, J. Am. Chem. Soc., 2024, 146, 18804–18810.
39. H. Park, K. M. Kim, A. Lee, S. Ham, W. Nam and J. Chin, Bioinspired Chemical Inversion of l-Amino Acids to d-Amino Acids, J. Am. Chem. Soc., 2007, 129, 1518–1519.
40. V. Babin, A. Talbot, A. Labiche, G. Destro, A. Del Vecchio, C. S. Elmore, F. Taran, A. Sallustrau and D. Audisio, Photochemical Strategy for Carbon Isotope Exchange with CO₂, ACS Catal., 2021, 11, 2968–2976.
41. D. Kong, M. Munch, Q. Qiqige, C. J. C. Cooze, B. H. Rotstein and R. J. Lundgren, Fast Carbon Isotope Exchange of Carboxylic Acids Enabled by Organic Photoredox Catalysis, J. Am. Chem. Soc., 2021, 143, 2200–2206.
42. (a) K. Donabauer, M. Maity, A. L. Berger, G. S. Huff, S. Crespi and B. König, Photocatalytic carbanion generation – benzylation of aliphatic aldehydes to secondary alcohols, Chem. Sci., 2019, 10, 5162–5166; (b) Q.-Y. Meng, T. E. Schirmer, A. L. Berger, K. Donabauer and B. König, Photocarboxylation of Benzylic C–H Bonds, J. Am. Chem. Soc., 2019, 141, 11393–11397; (c) S. Grotjahn and B. König, Photosubstitution in Dicyanobenzene-based Photocatalysts, Org. Lett., 2021, 23, 3146–3150.
43. J. Ning, B. Du, S. Cao, X. Liu and D. Kong, Combining Umpolung and Carbon Isotope Exchange Strategies for Accessing Isotopically Labeled α-Keto Acids, Org. Lett., 2024, 26, 5966–5971.
44. J. Ning, G. Jiang, H. Du and D. Kong, Direct Carbon Isotope Exchange of α-Keto Acids Enabled by Photoredox Self-Catalysis, Org. Lett., 2025, 27, 5995–6000.
45. For early examples on Henkel reaction, see: (a) B. Raecke, Angew. Chem., 1958, 70, 1–5; (b) E. McNelis, Reactions of Aromatic Carboxylates. II.¹ The Henkel Reaction, J. Org. Chem., 1965, 30, 1209–1213.
46. S. Furuyama, An Investigation of the Mechanism of the Henkel Reaction by Deuterium and Carbon-14 Isotope Techniques, Bull. Chem. Soc. Jpn., 1967, 40, 1212–1217.
47. (a) Y. Ogata, M. Tsuchida and A. Muramoto, The Preparation of Terephthalic Acid from Phthalic or Benzoic Acid, J. Am. Chem. Soc., 1957, 79, 6005–6008; (b) Y. Ogata, M. Hojo and M. Morikawa, Further Studies on the Preparation of Terephthalic Acid from Phthalic or Benzoic Acid, J. Org. Chem., 1960, 25, 2082–2087; (c) S. Thiyagarajan, A. Pukin, J. van Haveren, M. Lutz and D. S. van Es, Concurrent formation of furan-2,5- and furan-2,4-dicarboxylic acid: unexpected aspects of the Henkel reaction, RSC Adv., 2013, 3, 15678–15686.
48. (a) K. Fujishiro and S. Mitamura, The Zinc(II)-Catalyzed Henkel Reaction of Dipotassium 1,8-Naphthalenedicarboxylate in a Dispersion Medium, Bull. Chem. Soc. Jpn., 1989, 62, 786–790; (b) J. Szammer, E. Simon-Trompler and L. Ötvös, Catalysts of thermal exchange reactions between carboxyl groups and ¹⁴CO₂, J. Radioanal. Nucl. Chem., 1989, 135, 125–129.
49. L.-L. Liao, G.-M. Cao, Y.-X. Jiang, X.-H. Jin, X.-L. Hu, J. J. Chruma, G.-Q. Sun, Y.-Y. Gui and D.-G. Yu, α-Amino Acids and Peptides as Bifunctional Reagents: Carbocarboxylation of Activated Alkenes via Recycling CO₂, J. Am. Chem. Soc., 2021, 143, 2812–2821.
50. S. Wang, I. Larrosa, H. Yorimitsu and G. J. P. Perry, Carboxylic Acid Salts as Dual-Function Reagents for Carboxylation and Carbon Isotope Labeling, Angew. Chem., Int. Ed., 2023, 62, e202218371.
51. For experimental pK_a values see the tables made by F. G. Bordwell, D. H. Ripin and D. A. Evans via the ACS organic division website https://organicchemistrydata.org/.
52. X. Liu, S. Cao, C. Zhang, Y. Jiang and D. Kong, Integrating Umpolung and CO₂ Shuttling Strategies for the Synthesis of ¹²C- and ¹³C-α-Ketoacids from Aldehydes, Org. Lett., 2024, 26, 8967–8972.
53. K. Shimotai, O. Sasamoto and M. Shigeno, Carboxylations of (Hetero)Aromatic C–H Bonds Using an Alkyl Silyl Carbonate Reagent, Org. Lett., 2025, 27, 352–356.
54. (a) M. Shigeno, K. Hanasaka, K. Sasaki, K. Nozawa-Kumada and Y. Kondo, Direct Carboxylation of Electron-Rich Heteroarenes Promoted by LiO-tBu with CsF and [18]Crown-6, Chem. – Eur. J., 2019, 25, 3235–3239; (b) M. Shigeno, K. Hanasaka, I. Tohara, K. Izumi, H. Yamakoshi, E. Kwon, K. Nozawa-Kumada and Y. Kondo, Direct C–H Carboxylation Forming Polyfunctionalized Aromatic Carboxylic Acids by Combined Brønsted Bases, Org. Lett., 2022, 24, 809–814; (c) M. Shigeno, M. Kiriyama, K. Izumi, K. Sasaki, O. Sasamoto, K. Nozawa-Kumada and Y. Kondo, LiO-t-Bu/CsF-Mediated Formal Hydrolysis of Trifluoromethyl Arenes, Synthesis, 2024, 1438–1448; (d) M. Shigeno, K. Sasaki, K. Nozawa-Kumada and Y. Kondo, Double-Carboxylation of Two C–H Bonds in 2-Alkylheteroarenes Using LiO-t-Bu/CsF, Org. Lett., 2019, 21, 4515–4519; (e) M. Shigeno, I. Tohara, K. Nozawa-Kumada and Y. Kondo, Direct C-2 Carboxylation of 3-Substituted Indoles Using a Combined Brønsted Base Consisting of LiO-tBu/CsF/18-crown-6, Eur. J. Org. Chem., 2020, 1987–1991; (f) M. Shigeno, I. Tohara, K. Nozawa-Kumada and Y. Kondo, 1,5-Double-Carboxylation of 2-Alkylheteroarenes Mediated by a Combined Brønsted Base System, Synlett, 2022, 1376–1380; (g) M. Shigeno, I. Tohara, K. Sasaki, K. Nozawa-Kumada and Y. Kondo, Combined Brønsted Base-Promoted CO₂ Fixation into Benzylic C–H Bonds of Alkylarenes, Org. Lett., 2022, 24, 4825–4830.
55. X. Liu, G. J. P. Perry and D. Kong, A Para-Selective Kolbe–Schmitt Reaction, Angew. Chem., Int. Ed., 2025, 65, e22503.
56. K. E. Marris, J. T. Thorne, D. J. Ryder-Mahoney, R. A. Bragg, C. S. Elmore and G. J. P. Perry, Switchable N–H vs. C3−H carboxylation of indoles using dual-function reagents, Chem. Commun., 2026, 62, 6101–6105.
57. X. Liu, Z. Li, C. Song, G. Xu and D. Kong, Regiodivergent N1- and C3- carboxylation of indoles, Org. Chem. Front., 2026 10.1039/D6QO00306K.
58. D. J. Ryder-Mahoney, K. Yamazaki and G. J. P. Perry, Aryl Halide Carboxylation via Decarboxylative Metal-Halogen Exchange., JACS Au, 2026, 6, 773–778.
59. K. D. Collins and F. Glorius, A robustness screen for the rapid assessment of chemical reactions, Nat. Chem., 2013, 5, 597–601.
60. (a) G. J. P. Perry, J. M. Quibell, A. Panigrahi and I. Larrosa, Transition-Metal-Free Decarboxylative Iodination: New Routes for Decarboxylative Oxidative Cross-Couplings, J. Am. Chem. Soc., 2017, 139, 11527–11536; (b) J. M. Quibell, G. J. P. Perry, D. M. Cannas and I. Larrosa, Transition-metal-free decarboxylative bromination of aromatic carboxylic acids, Chem. Sci., 2018, 9, 3860–3865; (c) J. M. Quibell, G. Duan, G. J. P. Perry and I. Larrosa, Decarboxylative Suzuki–Miyaura coupling of (hetero)aromatic carboxylic acids using iodine as the terminal oxidant, Chem. Commun., 2019, 55, 6445–6448.

---

Footnote

† These authors contributed equally.

---