Skip to main content
NCBI home page
Nature Communications logoLink to Nature Communications
. 2026 Aug 25;17:10166. doi: 10.1038/s41467-026-77092-0

Amide synthesis via molecular shuffling of carboxylic acids

Chenglong Li 1, Qihang Tan 1, Qiang Wu 1, Yun Fu 1, Yongmei Wang 1, Bin Song 1, Tianzeng Huang 1, Long Liu 1, Michal Szostak 2,✉, Tieqiao Chen 1,✉
PMCID: PMC13601603  PMID: 42778545

Abstract

The amide bond is fundamental in chemistry and biology and its classical construction from carboxylic acids and amines is limited by positional connectivity of the carboxylic acid group. Herein, we report an approach to amide synthesis that relies on molecular shuffling of carboxylic acid group. In this reaction, amino benzoic acids engage in oxidative addition to a Pd(0) complex, and CO from decarbonylation inserts into the C–Pd bond formed by alkene insertion before the last reductive elimination, achieving carbonyl shuttle amidation. This amide synthesis is precisely controlled by ligand selection, where the reaction pathway can be fully switched to a divergent amine synthesis by decarbonylative amination. This method features wide substrate scope and excellent functional-group tolerance. Moreover, it is scalable and applicable to linear amides, highlighting its synthetic value. The study opens avenues for amide synthesis by carbonyl reshuffling and advances their use as critical building blocks in chemical science.

Subject terms: Synthetic chemistry methodology, Synthetic chemistry methodology, Homogeneous catalysis

The amide bond represents the most fundamental functional group in chemistry and biology. Herein, the authors report an approach to amide synthesis that relies on molecular shuffling of carboxylic acid functional group.

Introduction

The amide bond represents the most fundamental structural motif in chemistry and biology1. For decades, amides have been the most ubiquitous functional group in pharmaceuticals2,3, polymers4,5, materials science6,7 and other fields of science8–10. The synthesis of amides has been studied for over a century, with a key discovery of activated carboxylic acid derivatives by Sheehan in 195511, which enabled the synthesis of penicillin antibiotics (Fig. 1A)12,13. The most common synthetic route to amides involves condensation of such activated carboxylic acids with amines, while the importance of this pathway is highlighted by thousands of examples in the synthesis of pivotal amide bonds in all facets of chemical science14,15. Furthermore, this method has been widely exploited in intramolecular amidations, where macrocyclic amide linkages are key targets in drug discovery16–18, molecular receptors19–21 and conformational recognition22,23. After the studies by Sheehan, a significant collection of activating reagents has been established, including carbodiimides, uronium salts, triazines and phosphonium salts24–27, while more recent studies have focused on amidations using boron-based reagents28–31 and N-heterocyclic carbenes32. However, this classic approach is limited by the positional connectivity of the key carboxylic acid group. Herein, we report an approach to the synthesis of amides that relies on molecular shuffling of the carboxylic acid functional group (Fig. 1B). In this paradigm, amino benzoic acids engage in oxidative addition to a Pd(0) complex, and CO generated by decarbonylation undergoes transfer insertion into the C–Pd bond formed by alkene insertion before the last reductive elimination, achieving carbonyl shuttle amidation. The carbonyl shuttle amide synthesis is precisely controlled by ligand selection, where the reaction pathway can be fully switched to a divergent amine synthesis by decarbonylative amination. Moreover, this strategy is applicable to the synthesis of linear amides. The study opens avenues for the synthesis of amides by carbonyl reshuffling and advances an accelerated utilization of amide functional groups as critical building blocks in chemical science.

Fig. 1. Amide synthesis and transition-metal-catalyzed transformation of carboxylic acids.

Fig. 1

Open in a new tab

A Classical amide synthesis. B This study: amide synthesis by carboxylic acid reshuffling. C Blueprint for transition-metal shuttle catalysis. D General modes of transition-metal activation of carboxylic acids.

Recent years have witnessed the development of shuttle catalysis of carboxylic acid derivatives for organic synthesis (Fig. 1C)33–39. This approach capitalizes on the ubiquitous nature of carboxylic acids, where, due to the ready accessibility and key importance, developing highly efficient transformations of carboxylic acids facilitates the synthesis of target compounds and enables late modification strategies of common molecules bearing carboxylic acid functional groups. In this context, significant advances have been made in transition-metal-catalyzed activation of carboxylic acids during the past decade (Fig. 1D)40–50. Thus, carboxylic acids can serve as an acyl source to produce the target carboxylic derivatives43,44 or ketones45,46 or alternatively carboxylic acids can be engaged in decarboxylative couplings mediated by transition metals in the presence of oxidants47,48. Another pathway involves decarbonylative coupling of carboxylic acids, which has emerged as a powerful method for constructing chemical bonds49,50. In the process, carboxylic acids are in situ activated to form an active ester or acid derivative, and then act as powerful electrophiles to participate in diverse cross-couplings through decarbonylation under redox-neutral conditions. A wide range of chemical bonds has been efficiently accessed by this strategy.

In this work, we propose an approach to the construction of amides by merging the fields of shuttle catalysis and redox-neutral decarbonylative transformations. By carefully optimizing a bidentate phosphine/Pd system, CO generated via decarbonylation can efficiently re-insert into the C–Pd bond formed by alkene insertion prior to reductive elimination, thus enabling carbonyl shuttle amidation. This amide synthesis is fully catalyst-controlled, where electron-rich, monodentate phosphines, such as cataCXium A, result in decarbonylative amination, resulting in a divergent pathway to afford valuable amines. This general approach well overcomes the competing decarbonative side reactions during the reaction and represents successful carbonyl shuttle catalysis to form amide bonds. This approach is characterized by a particularly wide substrate scope and excellent functional group tolerance, including bioactive molecules and late-stage functionalization. The method is scalable, which highlights the practical utility of this method for the synthesis of amides. Extensive mechanistic studies provide key insights into this process. Overall, this pathway reforms the toolbox of amide bond synthesis by re-routing the classical carboxylic acid/amine condensation to carboxylic acid re-shuffling. Considering the significance of amides in pharmaceuticals, functional materials and polymers, this approach opens vistas for broad applications in chemical science.

Results

Reaction optimization

We initiated our investigation by examining the model reaction of 2-(phenylamino)benzoic acid 1a with norbornene 1b and the selected optimization results are summarized in Table 1. We found that in the presence of Pd(PPh3)4 (5 mol%) and bidentate phosphine ligand dppb (10 mol%), a mixture of 1a (0.2 mmol), 1b (4 equiv) and Piv2O (2 equiv) was heated in dioxane at 110 °C to produce the carboxylic acid reshuffled amide product in 87% yield with excellent selectivity (Table 1, Entry 1). The structure was confirmed by X-ray crystallographic analysis (Fig. 2A). As shown, the ORTEP drawing reveals that the product resulted from the CO shuttle of the carboxylic acid to the distal position of the amine. The reaction represents an example of the amide bond formation by CO shuttle catalysis of carboxylic acids.

Table 1.

Amide Synthesis by Carboxylic Acid Reshuffling.a,b

graphic file with name 41467_2026_77092_Taba_HTML.gif
Entry [Pd] (mol%) Ligand (mol%) Solvent (mL) 1c (%)b 1 d (%)b
1 Pd(PPh3)4 (5) dppb (10) dioxane (4) 87 trace
2c Pd(PPh3)4 (5) dppb (10) dioxane (4) 75 11
3d Pd(PPh3)4 (5) dppb (10) dioxane (4) 45 14
4e Pd(PPh3)4 (5) dppb (10) dioxane (4) 11 trace
5f Pd(PPh3)4 (5) dppb (10) dioxane (4) 95 (91 g) trace
6h Pd(OAc)2 (5) CataCXium A (20) dioxane (8) trace 96 (93 g)
7h Pd(OAc)2 (5) CataCXium A (20) dioxane (8) trace 79
8h Pd(OAc)2 (5) CataCXium A (20) dioxane (8) trace 64
9h Pd(OAc)2 (5) - dioxane (8) N.D. trace
10h Pd(OAc)2 (5) PCy (20) dioxane (8) 10 63
11h Pd(OAc)2 (5) PPh3 (20) dioxane (8) 20 9
12h Pd(OAc)2 (5) Xantphos (10) dioxane (8) 21 48
13h Pd(OAc)2 (5) dppe (10) dioxane (8) 5 24
14h Pd(OAc)2 (5) dppb (10) dioxane (8) 32 15
15h Pd(OAc)2 (5) dpph (10) dioxane (8) trace 8
16h Pd(OAc)2 (5) dppf (10) dioxane (8) 29 20
17h Pd(OAc)2 (5) dppm (10) dioxane (8) 14 8
18h Pd(OAc)2 (5) CataCXium A (20) NMP (8) trace trace
19h Pd(OAc)2 (5) CataCXium A (20) PhOMe (8) 17 27
20h Pd(OAc)2 (5) CataCXium A (20) toluene (8) 13 59
21h Pd(OAc)2 (5) CataCXium A (20) cyclohexane (8) 29 60
22i Pd(OAc)2 (5) CataCXium A (20) dioxane (8) N.D. trace
Open in a new tab

aReaction conditions: 1a (0.2 mmol), 1b (0.8 mmol), Pd(PPh3)4 (5 mol%), dppb (10 mol%), Piv2O (2 equiv), dioxane (4 mL), 110 °C, N2, 12 h.

bGC yield using tridecane as an internal standard.

c130 °C.

d140 °C.

e90 °C.

f24 h.

gIsolated yield.

h1.5 equiv Piv2O, 2 equiv 1b, 150 °C.

iWithout Piv2O, 2 equiv 1b, 150 °C. Piv2O: pivalic anhydride, NMP: N-Methylpyrrolidone, cataCXium A: di(1-adamantyl)-n-butylphosphine, N.D.: not detected.

Fig. 2. ORTEP drawing of compounds 1c & 1e.

Fig. 2

Open in a new tab

Thermal ellipsoids: 50% probability. A Selected bond lengths (Å) and angles (deg) of 1c; B Selected bond lengths (Å) and angles (deg) of 1e.

Several other optimization results are worth noting. First, we deduced that low reaction temperature could facilitate CO capture, thus favoring the formation of the amide. Indeed, when the reaction was conducted at a higher reaction temperature (130 °C or 140 °C), the yield of amide decreased (Table 1, Entries 2 and 3). However, further lowering the reaction temperature to 90 °C led to a decrease in the yield (Table 1, Entry 4). The result would be ascribed to a more difficult decarbonylation under these conditions. Finally, a very high yield of 1c was obtained by extending the reaction time to 24 h at 110 °C with full selectivity (Table 1, Entry 5). Second, there is a very significant ligand effect on the reaction, with the carboxylic acid reshuffling being completely controlled by the ancillary ligand. As such, we found that when electron-rich monodentate phosphine ligands such as cataCXium A (20 mol%) are used, the reaction after slightly tuning the reaction conditions is completely switched to a divergent decarbonylative amination to produce the corresponding decarbonylative cyclization product in 96% yield with full selectivity (Table 1, Entry 6). The phosphine ligand is essential to this reaction (Table 1, Entry 9). Other selected phosphine ligands, such as PCy3, PPh3, Xantphos, dppe, dppb, dpph, dppf and dppm, are less selective (Table 1, Entries 10–17). In terms of solvents, the reaction progressed slowly in NMP, PhOMe, toluene and cyclohexane (Table 1, Entries 18–21). The anhydride activator is essential (Table 1, Entry 22), as expected for the decarbonylative pathway. Finally, it should be noted that full selectivity for the formation of the amide reshuffling product is observed using the bidentate dppb ligand. In general, we also noticed that increasing olefin and activator concentration results in higher yields and selectivity (vide infra, mechanistic studies).

Substrate scope

With the optimized reaction conditions in hand, the substrate scope of this amide synthesis was investigated. As outlined in Fig. 3, this approach is broadly general and permits a synthesis of a wide range of amides by carbonyl reshuffling. Particularly worth noting is high functional group tolerance. Various valuable functional groups, such as alkyl, methoxyl, fluoro, chloro, silyl, boric, thiomethoxyl, trifluoromethyl, nitro, vinyl, phenyl, carbonyl, ester and heterocyclic groups are well tolerated under the present reaction conditions. High yields were achieved using both electron-rich and electron-deficient amines. For instance, derivatives bearing alkyl, methoxy, halo (F and Cl), carbonyl, CF3 and NO2 all showed high reactivity under the reaction conditions (Fig. 3, 2c–14c). Boric, thiomethyl, silyl, and vinyl groups were well compatible (Fig. 3, 15c–18c). It is worth noting that the products could be easily further utilized as functional handles through the transformation of these functional groups. π-Extended and heterocyclic N-Ar1 are also well-compatible in this reaction (Fig. 3, 19c–24c). It is known that alkyl groups increase the basicity of amines and have a negative effect on catalyst deactivation and side reactions. To our delight, substrates with various N-alkyl groups also produced the desired products in good yields (Fig. 3, 25c–27c). Moreover, various N-Ar2, including those with various sensitive functional groups, could all efficiently promote this amide synthesis (Fig. 3, 28c–41c). In addition, functionalized norbornenes were also suitable for this reaction and gave the CO shuttle amide transform in high yields (Fig. 3, 42c–44c).

Fig. 3. Substrate scope of amide synthesis by carboxylic acid reshuffling.

Fig. 3

Open in a new tab

Reaction conditions: benzoic acid a (0.2 mmol), alkene b (4 equiv), Pd(PPh3)4 (5 mol%), dppb (10 mol%), Piv2O (2 equiv), dioxane (4 mL), 110 °C, N2, 24 h. Isolated yields. Me methyl, OMe methoxy, tBu tert-butyl, Bpin pinacolato boron, TMS trimethylsilyl.

Crucially, this approach to amide synthesis by carboxylic acid reshuffling is readily applicable to the transformation of pharmaceuticals and bioactive carboxylic acids (Fig. 4). For example, mefenamic acid is used as an antipyretic, analgesic, and nonsteroidal anti-inflammatory drug in the clinic. This compound participated readily in the carboxylic acid reshuffling to afford the amide product in high yield (Fig. 4, 45c). Clofenamic acid, flufenamic acid and meclofenamic acid are also nonsteroidal anti-inflammatory drugs used to treat inflammation and as analgesics. These drugs are also suitable substrates for carboxylic acid reshuffling (Fig. 4, 46c–48c). The substrates bearing bioactive fragment eugenol and isoeugenol also showed high reactivity in the current catalytic system to furnish amides (Fig. 4, 49c, 50c). Interestingly, we found that this approach is also suitable for decarbonylative amine formation with a rather wide substrate scope (Fig. 5 and Supplementary Fig. S1).

Fig. 4. Modification of pharmaceuticals and bioactive molecules.

Fig. 4

Open in a new tab

Reaction conditions: benzoic acid a (0.2 mmol), NBE 1b (4 equiv), Pd(PPh3)4 (5 mol%), dppb (10 mol%), Piv2O (2 equiv), dioxane (4 mL), 110 °C, N2, 24 h. Isolated yields. NBE: norbornene.

Fig. 5. Decarbonylative amination.

Fig. 5

Open in a new tab

Conditions: benzoic acid a (0.2 mmol), alkene b (2 equiv), Pd(OAc)2 (5 mol%), cataCXium A (20 mol%), Piv2O (1.5 equiv), dioxane (8 mL), 150 °C, N2, 12 h. Isolated yields.

While our initial focus was on the reaction discovery of the carboxylic acid reshuffling for amide synthesis in the intramolecular context, we were pleased to find that this methodology is also applicable to linear amides (Figs. 2B and 6). It should be noted that X-ray analysis shows a trans-configuration on the norbornene unit. The trans/cis ratio is usually over 20:1. The high selectivity would be ascribed to the high temperature in this reaction35. This establishes the CO shuttle catalysis concept for the synthesis of acyclic amides by CO migration from common carboxylic acids.

Fig. 6. Substrate scope of linear amide synthesis by carboxylic acid reshuffling.

Fig. 6

Open in a new tab

Reaction conditions: acid a (0.2 mmol), NBE 1b (2.5 equiv), Et3N (3 equiv), Pd2(dba)3 (2.5 mol%), Xphos (10 mol%), (4-F-Ph)3P (1 mol%), Piv2O (1.5 equiv), Li2CO3 (1 equiv), solvent (cyclohexane: anisole = 1:1, 1 mL), 150 °C, N2, 10 h. Isolated yields. The trans/cis ratio was usually over 20:1. aPd(dba)2 (5 mol%) instead of Pd2(dba)3. bPd(acac)2 (5 mol%) instead of Pd2(dba)3. cTripropylamine instead of Et3N. dDMF instead of Et3N. eN-Methylformanilide instead of Et3N. Et3N: triethylamine, Xphos: 2-dicyclohexylphosphino-2′,4′,6′-triisopropylbiphenyl, DMF: N,N-dimethylformamide.

In practical terms, it is critical to point out that this carboxylic acid reshuffling is readily scalable (Fig. 7A). For example, the mixture of 1a (5 mmol) and 1b was heated under standard reaction conditions to afford the corresponding amide on a gram scale in 80% isolated yield. Similarly high yield was also achieved in decarbonylative amination on a gram scale.

Fig. 7. Synthetic applications and mechanistic studies.

Fig. 7

Open in a new tab

A Scale-up experiments. B Synthetic applications. C Control experiments. D CO control experiments. E Competition experiments of CO shuttle amidation. F Competition experiments of decarbonylative amination. iPr iso-propyl, Et ethyl, DIBAL-H diisobutylaluminium hydride.

Synthetic applications

As demonstrated in the substrate scope studies, this process for amide synthesis is distinguished by functional group tolerance to many valuable functional groups, facilitating further derivatization to afford target amide products that are not easily obtained by other methods. Furthermore, the amide itself provides a highly valuable handle for streamlined synthesis as a versatile synthetic intermediate (Fig. 7B). Taking 1c as an example, the amide bond can be easily reduced with LiAlH4 to produce the corresponding six-membered cyclic amine g in 77% yield. Likewise, the amide bond can efficiently react with organometallic reagents, such as MeLi, to generate cyclic amine h. Furthermore, Ti(OiPr)4-mediated Kulikovich cyclopropanation with EtMgBr delivers the spiro amine i. Finally, treatment with DIBAL-H and dehydration furnishes the active imine salt j in high yield. These reactions further highlight the synthetic value of the present carboxylic acid reshuffling amide bond synthesis. Furthermore, quinoline and indoline heterocycles have featured many prominent applications in drug discovery and materials science.

Mechanistic investigations

To gain mechanistic insights into this reaction, extensive studies were conducted. First, we synthesized the mixed anhydride la and found that it could be converted into 1c in 91% yield and into 1d in 94% yield, indicating that the mixed anhydride is the active intermediate (Fig. 7C, left). As mentioned, the N–H moiety can be acylated by anhydride to produce f as detected by GC-MS. To verify whether the N-Piv product could participate in the reaction, we synthesized this potential intermediate and subjected it to the reaction conditions. However, when f was reacted under the respective conditions, only a trace amount of 1d was generated, while 1c was not detected (Fig. 7C, right). These results suggest that f is an unproductive byproduct of this process.

In the CO shuttle amidation, CO was re-utilized with very high efficiency. Thus, we carried out a reaction in glass tubes with different volumes (Fig. 7D, left). The yields in 10 mL, 25 mL, and 100 mL reactors did not change significantly and were 94%, 93% and 82%, respectively. We also conducted the reaction with different loading of CO adsorbent and found that the yields also varied from 89–91% (Fig. 7D, right). These results indicate that CO does not permeate outside the reaction system and coordinates with palladium during the reaction, thus enabling the successful utilization of carboxylic acid reshuffling.

To gain insight into the selectivity of this process, intermolecular competition experiments were conducted (Fig. 7E, F). When 8a and 1a were allowed to react with NBE under carboxylic acid reshuffling conditions, 7c and 1c were produced in 14% and 25% yields, respectively (Eq 1). When 20a competed with 1a, 13c was obtained in 19% yield, while 1c was generated in 20% yield (Eq 2). The competition between 8a and 20a produced 7c and 13c in 15% and 22% yields (Eq 3). Similar results were obtained for decarbonylative amination (Eqs 7–9). These results indicate that although the electronic effect of N-Ar1 is not significant, substrates with electron-deficient N-aryl1 are inherently favored. Meanwhile, allowing 34a to compete with 1a under carboxylic acid reshuffling conditions produced 30c and 1c in 14% and 30% yields, respectively (Eq 4). When 38a competed with 1a, 34c was obtained in 35% yield, while 1c was produced in 11% yield (Eq 5). The competition of 34a with 38a also generated 30c and 34c in 4% and 20% yields (Eq 6). Similar results were obtained for decarbonylative amination (Eqs 10–12). These results indicate that electron-deficient benzoic acids are inherently favored, which was also supported by Hammett analysis. It is deduced that the electron-deficient benzoic acids would facilitate the formation of mixed anhydrides, oxidative addition and decarbonylation, thus promoting the reaction51,52.

The Hammett analysis was subsequently conducted53. A positive slope (ρ = 0.259) was observed for the CO shuttle amidation of benzoic acids having para-substituted N-Ar with NBE 1b (Fig. 8a). A larger value (ρ = 1.909) was obtained for the same reaction of benzoic acids bearing para-substituted benzene ring with 1b (Fig. 8b). In the decarbonylative amination, similar slope values were observed (Fig. 8c, d). The data indicate that electron-deficient substrates are inherently more favorable. Compared with N-Ar, the electronic effect at the benzene ring of benzoic acids is comparatively larger and thus contributes more significantly to the rate-determining step. The Hammett plot results are consistent with intermolecular competition experiments.

Fig. 8. Hammett effects and kinetic analysis.

Fig. 8

Open in a new tab

Hammett plots of CO shuttle amidation: a Benzoic acids having para-substituted N-Ar with NBE 1b; b Benzoic acids bearing para-substituted benzene ring with NBE 1b; hammett plots of decarbonylative amination: c Benzoic acids having para-substituted N-Ar with NBE 1b; d Benzoic acids bearing para-substituted benzene ring with NBE 1b; kinetic analysis of CO shuttle amidation: e 2-(Phenylamino)benzoic acid vs initial rate, f NBE vs initial rate, g [Pd] vs initial rate, h Eyring plot: plot of ln k vs 1/T; kinetic analysis of decarbonylative amination: i 2-(phenylamino)benzoic acid vs initial rate, j NBE vs initial rate, k [Pd] vs initial rate, l Eyring plot: plot of ln k vs 1/T. Ea the activation energy, ΔH≠ the enthalpy of activation, ΔS≠ the entropy of activation, ΔG≠ the Gibbs energy of activation, k the rate constant, A the pre-exponential factor, T temperature, R gas constant (8.314 J/molK).

Next, the kinetic analysis was performed for the CO shuttle amidation and showed that the rate was first-order dependent on benzoic acids, palladium and olefins, indicating that norbornenes had a significant effect on the rate (Fig. 8e–g and Supplementary Figs. S2–S4). The activation energy was calculated as 26.43 kcal/mol based on the temperature effect (Fig. 8h and Supplementary Fig. S5)54. However, in the case of decarbonylative amination, the outcome was different. The rate was first-order dependent on benzoic acids and palladium catalyst, but close to zero-order dependent on norbornenes (Fig. 8i–k and Supplementary Figs. S6–S8). These results suggest that norbornenes do not have a major effect on the rate during the reaction and that the process is a mono-palladium catalytic process. The activation energy also increased (30.35 kcal/mol), which is consistent with the comparatively higher temperature employed in this reaction (Fig. 8l and Supplementary Fig. S9).

On the basis of control experiments, competition studies, Hammett analysis and kinetic studies, a plausible mechanism is proposed as shown in Fig. 9. Benzoic acids are first activated to give the active mixed anhydrides, which undergo oxidative addition of Pd(0) into the C(O)-acyl bond to produce the acyl palladium species II. The resulting intermediate undergoes decarbonylation, alkene insertion, intramolecular ligand exchange and reductive elimination via a direct decarbonylative pathway. However, when the ligand is switched from cataCXium A to bidentate dppp, a transient decarbonylation of species Ⅱ takes place to afford the CO-ligated aryl palladium Ⅲ. This intermediate undergoes alkene insertion, intramolecular ligand exchange, and key CO re-insertion to generate amides after reductive elimination. This cycle represents an approach to amide synthesis, where the carboxylic acid is positionally reshuffled to form N–C(O) bonds in analogy to the venerable acid/amine condensation. Mechanistically, the high reactivity of norbornenes can be ascribed to the ring strain of bicyclic alkenes. It is worth noting that norbornene scaffolds are highly valuable in chemistry as rigid, three-dimensional bicyclic frameworks in cancer drug design, structural scaffolding, polymer chemistry and materials science.

Fig. 9. Proposed catalytic cycle.

Fig. 9

Open in a new tab

Proposed mechanism of carboxylic acid reshuffling (left); proposed mechanism of decarbonylative amination (right).

Discussion

In conclusion, amide bonds have been at the heart of organic synthesis for more than a century. The classical and most widely utilized method for amide synthesis involves the condensation of activated carboxylic acids with amines, and this approach has been exploited in every facet of chemical science on a daily basis. Herein, we developed a paradigm for the synthesis of amides that capitalizes on the molecular reshuffling of the carboxylic acid functional group. In this blueprint, amino benzoic acids engage in oxidative addition to a low valent metal, and CO generated by decarbonylation undergoes transfer insertion into alkenes before the last reductive elimination. This general approach overcomes the competing decarbonative side reactions during the reaction and represents the successful carbonyl shuttle catalysis to form amide bonds. This carbonyl shuttle amide synthesis is precisely controlled by ligand selection, where the reaction pathway can be fully switched to a divergent amine synthesis by decarbonylative amination. This transform is capable of forming amides by the repositioned classical carboxylic acid/amine condensation and should greatly advance vistas on the synthesis of valuable amide bonds by carbonyl reshuffling.

Methods

General procedure for the synthesis of amides via carboxylic acid reshuffling

An oven-dried 25 mL Schlenk tube was charged with 2-aminobenzoic acid a (0.2 mmol, 1 equiv), NBE b (0.8 mmol, 4 equiv), Pd(PPh3)4 (5 mol%), and dppb (10 mol%). Subsequently, Piv2O (0.4 mmol, 2 equiv) and dioxane (4 mL) were added under N2. The reaction mixture was heated at 110 °C for 24 h. After completion of the reaction, the reaction mixture was cooled to room temperature. The residue was purified by flash column chromatography on silica gel (petroleum ether/EtOAc = 5:1) to afford the corresponding products 1c–50c.

General procedure for the synthesis of amines by decarbonylative amination

An oven-dried 25 mL Schlenk tube was charged with 2-aminobenzoic acid a (0.2 mmol, 1 equiv), NBE b (0.4 mmol, 2 equiv), Pd(OAc)2 (5 mol%), and cataCXium A (20 mol%). Subsequently, Piv2O (0.3 mmol, 1.5 equiv) and dioxane (8 mL) were added under N2. The reaction mixture was heated at 150 °C for 12 h. After completion of the reaction, the reaction mixture was cooled to room temperature. The residue was purified by flash column chromatography on silica gel (petroleum ether) to afford the corresponding products 1d–44d.

General procedure for the synthesis of linear amides

An oven-dried 25 mL Schlenk tube was charged with carboxylic acid a (0.2 mmol, 1 equiv), NBE 1b (0.5 mmol, 2.5 equiv), Et3N (0.6 mmol, 3 equiv), Pd2(dba)3 (5 mol%), Xphos (10 mol%), tris(4-fluorophenyl)phosphine ((4-F-Ph)3P, 1 mol%) and Li2CO3 (1 equiv). Subsequently, Piv2O (0.3 mmol, 1.5 equiv), cyclohexane/anisole (1:1, 1 mL) were added under N2. The reaction mixture was heated at 150 °C for 10 h. After completion of the reaction, the reaction mixture was cooled to room temperature. The residue was purified by flash column chromatography on silica gel to afford the corresponding products 1e–65e.

Supplementary information

Supplementary Information (28.8MB, pdf)
Transparent Peer Review file (89.6KB, pdf)

Acknowledgments

T.C. thanks the National Natural Science Foundation of China (Grant Nos. 22261015, 21871070) and the Key R&D project of Hainan province (No. ZDYF2020168, XTCX2022STA01) for financial support. M.S. thanks Rutgers University and the NSF (CAREER CHE-1650766) for financial support.

Author contributions

M.S. and T.C. conceived and designed the project. C.L., Q.T., Q.W., Y.F., and Y.W. developed the methodology. C.L. performed the investigation. T.C. acquired the funding. M.S. and T.C. supervised the project. M.S. and T.C. wrote the original draft. C.L., Q.T., B.S., T.H., L.L., M.S., and T.C. reviewed and edited the manuscript.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Funding

T. C. discloses support for the research of this work from the National Natural Science Foundation of China (Grant Nos. 22261015, 21871070) and the Key R&D project of Hainan province (Nos. ZDYF2020168, XTCX2022STA01). M. S. discloses support for the research of this work from Rutgers University and the NSF (Grant No. CAREER CHE-1650766).

Data availability

The X-ray crystallographic coordinates for structures reported in this study have been deposited at the Cambridge Crystallographic Data Center (CCDC), under deposition numbers 2530741 (1c) and 2530742 (1e). These data can be obtained free of charge from The Cambridge Crystallographic Data Center via www.ccdc.cam.ac.uk/data_request/cif. All other data supporting the findings of this study are available within the paper and its Supplementary Information. All data are also available from the corresponding author upon request.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

Michal Szostak, Email: michal.szostak@rutgers.edu.

Tieqiao Chen, Email: chentieqiao@hnu.edu.cn.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41467-026-77092-0.

References

  • 1.Greenberg, A., Breneman, C. M. & Liebman, J. F. The Amide Linkage: Structural Significance in Chemistry, Biochemistry and Materials Science. (Wiley, New York, 2000).
  • 2.Brown, D. G. & Boström, J. Analysis of past and present synthetic methodologies on medicinal chemistry: Where have all the new reactions gone? J. Med. Chem.59, 4443–4458 (2016). [DOI] [PubMed] [Google Scholar]
  • 3.For the top 200 small molecule pharmaceuticals by retail sales in 2023, see:https://bpb-us-e2.wpmucdn.com/sites.arizona.edu/dist/9/130/files/2024/07/2023Top200SmallMoleculePosterV6.pdf
  • 4.Goodridge, R. D., Tuck, C. J. & Hague, R. J. M. Laser sintering of polyamides and other polymers. Prog. Mater. Sci.57, 229–267 (2012). [Google Scholar]
  • 5.Winnacker, M. Terpene-based polyamides: a sustainable polymer class with huge potential. Curr. Opin. Green Sustain. Chem.41, 100819–100827 (2023). [Google Scholar]
  • 6.Shu, J., Wang, J. M., Li, Z. & Tong, K. Y. Effects of slit edge notches on mechanical properties of 3D-printed PA12 nylon kirigami specimens. Polymers15, 3082 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Hajibeygi, M., Shabanian, M. & Khodaei-Tehrani, M. New heat-resistant nanocomposites reinforced silicate nanolayers containing triazine rings based on polyamide: synthesis, characterization, and flame retardancy study. Polym. Compos.37, 188–198 (2016). [Google Scholar]
  • 8.Hughes, A. B. Ed, Amino acids, peptides and proteins in organic chemistry (Wiley, Weinheim, 2011).
  • 9.Crespo, L. et al. Peptide and amide bond-containing dendrimers. Chem. Rev.105, 1663–1681 (2005). [DOI] [PubMed] [Google Scholar]
  • 10.Dydio, P., Lichosyt, D. & Jurczak, J. Amide- and urea-functionalized pyrroles and benzopyrroles as synthetic, neutral anion receptors. Chem. Soc. Rev.40, 2971–2985 (2011). [DOI] [PubMed] [Google Scholar]
  • 11.Sheehan, J. C. & Hess, G. P. A new method of forming peptide bonds. J. Am. Chem. Soc.77, 1067–1068 (1955). [Google Scholar]
  • 12.Sheehan, J. C. & Henery-Logan, K. R. The total synthesis of penicillin V. J. Am. Chem. Soc.79, 1262–1263 (1957). [Google Scholar]
  • 13.Sheehan, J. C. & Henery-Logan, K. R. The total and partial general syntheses of the penicillins. J. Am. Chem. Soc.84, 2983–2990 (1962). [Google Scholar]
  • 14.March, J. In advanced organic chemistry. (John Wiley and Sons, New York, ed. 5, 2001).
  • 15.Trost, B. M. Editor, Comprehensive organic synthesis. (Pergamon Press, Oxford, 1991).
  • 16.White, C. J. & Yudin, A. K. Contemporary strategies for peptide macrocyclization. Nat. Chem.3, 509–524 (2011). [DOI] [PubMed] [Google Scholar]
  • 17.Mulligan, V. K. The emerging role of computational design in peptide macrocycle drug discovery. Expert Opin. Drug Discov.15, 833–851 (2020). [DOI] [PubMed] [Google Scholar]
  • 18.Jin, X., Ding, N., Guo, H. Y. & Hu, Q. Macrocyclic-based strategy in drug design: From lab to the clinic. Eur. J. Med. Chem.277, 116733 (2024). [DOI] [PubMed] [Google Scholar]
  • 19.Pan, Y. C., Tian, J. H. & Guo, D. S. Molecular recognition with macrocyclic receptors for application in precision medicine. Acc. Chem. Res.56, 3626–3639 (2023). [DOI] [PubMed] [Google Scholar]
  • 20.He, Q., Vargas-Zúñiga, G. I., Kim, S. H., Kim, S. K. & Sessler, J. L. Macrocycles as ion pair receptors. Chem. Rev.119, 9753–9835 (2019). [DOI] [PubMed] [Google Scholar]
  • 21.Pinalli, R., Pedrini, A. & Dalcanale, E. Biochemical sensing with macrocyclic receptors. Chem. Soc. Rev.47, 7006–7026 (2018). [DOI] [PubMed] [Google Scholar]
  • 22.Caricato, M. et al. Synthesis and anion recognition properties of shape-persistent binaphthyl-containing chiral macrocyclic amides. Beilstein J. Org. Chem.8, 967–976 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Chmielewski, M. J. & Jurczak, J. Anion recognition by neutral macrocyclic amides. Chem. Eur. J.11, 6080–6094 (2005). [DOI] [PubMed] [Google Scholar]
  • 24.Han, S. Y. & Kim, Y. A. Recent development of peptide coupling reagents in organic synthesis. Tetrahedron60, 2447–2467 (2004). [Google Scholar]
  • 25.Montalbetti, C. & Falque, V. Amide bond formation and peptide coupling. Tetrahedron61, 10827–10852 (2005). [Google Scholar]
  • 26.Valeur, E. & Bradley, M. Amide bond formation: beyond the myth of coupling reagents. Chem. Soc. Rev.38, 606–631 (2009). [DOI] [PubMed] [Google Scholar]
  • 27.El-Faham, A. & Albericio, F. Peptide coupling reagents, more than a letter soup. Chem. Rev.111, 6557–6602 (2011). [DOI] [PubMed] [Google Scholar]
  • 28.Ishihara, K., Ohara, S. & Yamamoto, H. 3,4,5-Trifluorobenzeneboronic acid as an extremely active amidation catalyst. J. Org. Chem.61, 4196–4197 (1996). [DOI] [PubMed] [Google Scholar]
  • 29.Tang, P. Organic syntheses. (John Wiley & Sons, Ltd, 2005).
  • 30.Al-Zoubi, R. M., Marion, O. & Hall, D. G. Direct and waste-free amidations and cycloadditions by organocatalytic activation of carboxylic acids at room temperature. Angew. Chem. Int. Ed.47, 2876–2879 (2008). [DOI] [PubMed] [Google Scholar]
  • 31.Charville, H. et al. The thermal and boron-catalyzed direct amide formation reactions: mechanistically understudied yet important processes. Chem. Commun.46, 1813–1823 (2010). [DOI] [PubMed] [Google Scholar]
  • 32.Qin, Y. L. & Li, T. T. N-heterocyclic carbene catalyzed activation of carboxylic acids. Asian J. Org. Chem.14, e202400585 (2024). [Google Scholar]
  • 33.Lee, Y. H., Denton, E. H. & Morandi, B. Modular cyclopentenone synthesis through the catalytic molecular shuffling of unsaturated acid chlorides and alkynes. J. Am. Chem. Soc.142, 20948–20955 (2020). [DOI] [PubMed] [Google Scholar]
  • 34.Lee, Y. H., Denton, E. H. & Morandi, B. Palladium-catalysed carboformylation of alkynes using acid chlorides as a dual carbon monoxide and carbon source. Nat. Chem.13, 123–130 (2021). [DOI] [PubMed] [Google Scholar]
  • 35.Denton, E. H. et al. Catalytic carbochlorocarbonylation of unsaturated hydrocarbons via C−COCl bond cleavage. Angew. Chem. Int. Ed.60, 23435–23443 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Ito, Y. et al. Nickel-catalyzed addition of C−C bonds of amides to strained alkenes: the 1,2-carboaminocarbonylation reaction. J. Am. Chem. Soc.144, 662–666 (2022). [DOI] [PubMed] [Google Scholar]
  • 37.Denton, E. H. et al. Ligand design enables the palladium-catalyzed intermolecular carbochlorocarbonylation of alkynes and cyclopentenone formation. J. Am. Chem. Soc.147, 25215–25223 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Zhang, R., Yu, T. & Dong, G. Rhodium-catalyzed tunable amide homologation through a hook-and-slide strategy. Science382, 951–957 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Kaldre, D., Klose, I. & Maulide, N. Stereodivergent synthesis of 1,4-dicarbonyls by traceless charge-accelerated sulfonium rearrangement. Science361, 664–667 (2018). [DOI] [PubMed] [Google Scholar]
  • 40.Varenikov, A., Shapiro, E. & Gandelman, M. Decarboxylative halogenation of organic compounds. Chem. Rev.121, 412–484 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Kitcatt, D. M., Nicolle, S. & Lee, A. Direct decarboxylative Giese reactions. Chem. Soc. Rev.51, 1415–1453 (2022). [DOI] [PubMed] [Google Scholar]
  • 42.Gooßen, L. J., Rodriguez, N. & Gooßen, K. Carboxylic acids as substrates in homogeneous catalysis. Angew. Chem. Int. Ed.47, 3100–3120 (2008). [DOI] [PubMed] [Google Scholar]
  • 43.Wang, S. P., Cheung, C. W. & Ma, J. A. Direct amidation of carboxylic acids with nitroarenes. J. Org. Chem.84, 13922–13934 (2019). [DOI] [PubMed] [Google Scholar]
  • 44.Tatsumi, K. et al. Zinc-catalyzed synthesis of acylsilanes using carboxylic acids and a silylborane in the presence of pivalic anhydride. Org. Lett.21, 10130–10133 (2019). [DOI] [PubMed] [Google Scholar]
  • 45.Shu, X. M. et al. Direct enantioselective C(sp3)−H acylation for the synthesis of α‑amino ketones. J. Am. Chem. Soc.142, 19058–19064 (2020). [DOI] [PubMed] [Google Scholar]
  • 46.Jiang, X. L. et al. Ligand relay catalysis enables asymmetric migratory reductive acylation of olefins or alkyl halides. J. Am. Chem. Soc.144, 21448–21456 (2022). [DOI] [PubMed] [Google Scholar]
  • 47.Zuo, Z. W. et al. Merging photoredox with nickel catalysis: coupling of α-carboxyl sp3-carbons with aryl halides. Science345, 437–440 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Hioki, Y. et al. Overcoming the limitations of Kolbe coupling with waveform-controlled electrosynthesis. Science380, 81–87 (2023). [DOI] [PubMed] [Google Scholar]
  • 49.Lu, H. et al. Selective decarbonylation via transition-metal-catalyzed carbon-carbon bond cleavage. Chem. Rev.121, 365–411 (2021). [DOI] [PubMed] [Google Scholar]
  • 50.Boehm, P. et al. Palladium-catalyzed decarbonylative iodination of aryl carboxylic acids enabled by ligand-assisted halide exchange. Angew. Chem. Int. Ed.60, 17211–17217 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Zhao, R. Y. et al. Thermoplastic high-strain multishape memory polymer: side-chain polynorbornene with columnar liquid crystalline phase. Adv. Mater.29, 1605908 (2017). [DOI] [PubMed] [Google Scholar]
  • 52.Xie, S. W. et al. Discovery of norbornene as a novel hydrophobic tag applied in protein degradation. Angew. Chem. Int. Ed.62, e202217246 (2023). [DOI] [PubMed] [Google Scholar]
  • 53.Hansch, C., Leo, A. & Taft, R. W. A survey of Hammett substituent constants and resonance and field parameters. Chem. Rev.91, 165–195 (1991). [Google Scholar]
  • 54.Nakamura, K. & Takayanagi, T. A modified Arrhenius equation. Chem. Phys. Lett.160, 295–298 (1989). [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Information (28.8MB, pdf)
Transparent Peer Review file (89.6KB, pdf)

Data Availability Statement

The X-ray crystallographic coordinates for structures reported in this study have been deposited at the Cambridge Crystallographic Data Center (CCDC), under deposition numbers 2530741 (1c) and 2530742 (1e). These data can be obtained free of charge from The Cambridge Crystallographic Data Center via www.ccdc.cam.ac.uk/data_request/cif. All other data supporting the findings of this study are available within the paper and its Supplementary Information. All data are also available from the corresponding author upon request.

Articles from Nature Communications are provided here courtesy of Nature Publishing Group

RESOURCES