A one-pot route to thioamides discovered by gas-phase studies: palladium-mediated CO₂ extrusion followed by insertion of isothiocyanates

Abstract

A new palladium mediated “one pot” synthesis of thioamides from aromatic carboxylic acids (ArCO₂H + RNCS → ArC(S)NHR + CO₂) was discovered by gas-phase experiments and theoretical studies. Palladium-mediated decarboxylation of aromatic carboxylic acids followed by addition of substituted isothiocyanates leads to the corresponding thioamide derivatives.

---

A desire to prepare new and known materials by efficient chemical synthesis using methods that are as environmentally benign as possible stimulates continued efforts to discover new reactions.^1–3 Computational chemistry is often used to probe the mechanisms of new reactions⁴ and mass spectrometry (MS) is often used to characterise products⁵ but rarely have these techniques been used together as tools for the discovery of new reactions.⁶ Here, we report their use to direct the hypothesis driven discovery of a new method to synthesise thioamides, an important class of compounds with applications in medicinal chemistry.⁷

The method involves ‘swapping’ the carboxylate functional group of readily available aromatic carboxylic acids (with the loss of CO₂) for a heterocumulene (eqn (1)). This new method could prove to be an attractive alternative to conventional methods that require multiple steps and/or harsh reaction conditions.⁸

ArCO₂H + RNCS → ArC(S)NHR + CO₂ (1)

Density Functional Theory (DFT) calculations suggested reactions represented by (1) to be exothermic for typical substrates,⁹ and this led us to investigate palladium mediated decarboxylative transformation of aromatic carboxylic acids into thioamides (Scheme 1). Many of the individual steps of the reaction scheme have precedence but they have not been used together to achieve a ‘one pot’ synthesis of thioamides from carboxylic acids.^10–15

Scheme 1 Key reactions of relevance to palladium catalysed decarboxylative transformation of aromatic carboxylic acids into thioamides (eqn (1)): the decarboxylative Heck reaction 1 → 2 → 3 → 4 (Myers);¹² protodecarboxylation 1 → 2 → 3 → 5 (Kozlowski);¹³ and thioamide synthesis 1 → 2 → 3 → 6 (this work) share the same steps for the formation of the organopalladium intermediate (3), but differ in its subsequent reactions.

To model the key decarboxylation and insertion steps associated with transformation of coordinated benzoate to coordinated thiobenzamides shown in Scheme 1, the two DMSO ligands were replaced by the phen ligand and loss of the anionic group X provided the requisite charge for the MS studies. Thus eqn (3) and (4) below were investigated using multistage mass spectrometry (MS) experiments and DFT calculations (Fig. 1).

[(L)Pd(X)₂] + ArCO₂H → [(L)Pd(X)(O₂CAr)] + HX (2)

[(phen)Pd(O₂CPh)]⁺ → [(phen)Pd(Ph)]⁺ + CO₂ (3)

[(phen)Pd(Ph)]⁺ + RNCS → [(phen)Pd(SNRCPh)]⁺ (4)

The complex [(phen)Pd(O₂CC₆H₅)]⁺ (phen = 1,10-phenanthroline) can be generated by electrospray ionization (ESI) and decarboxylated upon collision-induced dissociation (CID) to yield [(phen)Pd(C₆H₅)]⁺. Decarboxylation proceeds viaTS_7–8, transforming the chelating benzoate, 7, to reactive conformation 8, that yields [(phen)Pd(C₆H₅)(OCO)]⁺9 through TS_8–9 (Fig. 1B). Loss of CO₂ gives the putative three-coordinate complex [(phen)Pd(C₆H₅)]⁺10,¹⁶ which readily reacts with MeNCS (Fig. 1C and eqn (4)) to form [(phen)Pd(SC(NMe)C₆H₅)]⁺ at a rate of 2.03 cm³ molecules⁻¹ s⁻¹. The identity of the C–C bond coupled product was confirmed as [(phen)Pd(SC(NMe)C₆H₅)]⁺, rather than [(phen)Pd(C₆H₅)(MeNCS)]⁺, by comparing the CID spectra of the ion–molecule reaction product with the CID spectra of an authentic sample of the Pd²⁺ complex of N-methylbenzothioamide prepared by a standard Grignard reaction (Fig. S1, ESI†). A similar reaction (rate of 2.99 cm³ molecules⁻¹ s⁻¹) is observed for PhNCS. Both reactions are highly efficient, proceeding at the collision rate. DFT calculations predict that these reactions are highly exothermic, and occur through isothiocyanate insertion into the Pd–C bond to give the coordinated thioamides, [(phen)Pd(SC(NR)C₆H₅)]⁺ (Fig. 1D and Fig. S2, ESI†). DFT calculations also predict these reactions could be considered the reverse of decarboxylation (Fig. 1B), initially forming [(phen)Pd(C₆H₅)(SCNR)]⁺, 11, which then undergoes insertion viaTS_11–12 to give S-coordinated thioamide, 12, which isomerises to 13viaTS_12–13 (Fig. 1D).

Fig. 1 Gas-phase experiments (LTQ ion trap) and DFT calculations (B3LYP-gd3bj/SDD6-31+G(d)//M06/SDD6-31+G(d)) on: (A and B) decarboxylation (eqn (3), ΔG = 10.8 kcal mol⁻¹, ΔH = 23.7 kcal mol⁻¹, MS² CID, normalised collision energy of 14% and reaction time of 10 ms); (C and D) insertion of MeNCS into the Pd–C bond (eqn (4), MS³ IMR, concentration of MeNCS is 2.8 × 10⁹ molecule cm⁻³ and reaction time is 250 ms). DFT calculated species are orientated to highlight the direct relationship between carbon dioxide extrusion and isothiocyanate insertion for isoelectronic CO₂ and SCNR. Relative Gibbs and enthalpy energies (in parentheses, in kcal mol⁻¹). For (B) and (D) carbon = grey, nitrogen = blue, oxygen = red, palladium = turquoise, sulphur = yellow.

Stimulated by the proof of concept gas-phase studies, we explored the conversion of aromatic carboxylic acids into thioamides using a stoichiometric amount of palladium acetate. Individual reaction steps shown in Scheme 1 (eqn (2)–(5)) were examined via¹H NMR spectroscopy. Successful conversion relies on the insertion reaction (eqn (4)) being faster than protonation of the organometallic (eqn (6)) that leads to protodecarboxylation (eqn (7)).

[(L)Pd(X)(SNRCAr)] + “H source” → ArC(S)NHR (5)

[(L)Pd(X)(Ar)] + HX → [(L)Pd(X)₂] + ArH (6)

ArCO₂H → ArH + CO₂ (7)

Guided by Kozlowski's detailed mechanistic study on protodecarboxylation,¹³¹H NMR was used to monitor the transformation of 2,6-dimethoxybenzoic acid to N-alkyl-2,6-dimethoxythiobenzamide using Pd(O₂CCH₃)₂ in d₆-DMSO (Fig. S3–S7, ESI†). Formation of [(DMSO)_nPd(O₂CCH₃)(O₂CAr)] results in an upfield shift of the resonances attributed to the para-proton from δ 7.30 to 7.18 ppm (Fig. S4b, ESI†). Heating the reaction mixture at 65 °C for 4 hours results in decarboxylation and the formation of the organometallic complex [(DMSO)₂Pd(O₂CCH₃)(Ar)] (Fig. S4c (ESI†) and eqn (3)), and results in a further upfield shift of H_para to 7.01 ppm with concomitant peak broadening. Continued heating for 24 hours leads to formation of 1,3-dimethoxybenzene (Fig. S4d (ESI†) and eqn (6)), corresponding to the last step of the undesired protodecarboxylation reaction (eqn (7)). In contrast, if RNCS is added at room temperature to [(DMSO)₂Pd(O₂CCH₃)(Ar)], the coordinated thioamidate is formed (Fig. S4e (ESI†) and eqn (4)). The resultant NMR spectrum is complex, likely due to different coordination modes of the thioamidate to the Pd centre. The free thioamide is formed following reaction with NaBH₄ (Fig. S4f (ESI†) and eqn (5)). The overall process of insertion followed by reaction with NaBH₄ results in a downfield shift of the H_para proton (7.01 to 7.24 ppm, Δδ = 0.23 ppm) indicating the transformation to the thioamide. The identity of the thioamide was also confirmed by ESI-MS analysis of the NMR sample. Insertion is sensitive to the nature of the alkyl group of the isothiocyanate, RNCS (Fig. S4–S7, ESI†). Conversion to the thioamide occurs at 1 hour for R = Me, Et and iPr, but requires 4 hours for R = tBu. Relative yields for thioamide formation versus protodecarboxylation were: ≈80% for R = Me, Et and iPr and ≈ 50% for R = tBu, consistent with the longer reaction time required when R = tBu.

The effect of replacing the phen ligand used in the MS experiments with coordinated DMSO was investigated by DFT calculations.¹⁷ It is possible for the [Pd(DMSO)₂] complexes to exists as either cis or trans isomers (Fig. S41 and 44, ESI†). The presence of a ‘reservoir’ of acetic acid in solution (absent in the gas-phase) can lead to the formation of 1,3-dimethoxybenzene so competition between insertion and protonation of the organopalladium was considered. The solvation effect of DMSO on the optimised structures required the use of a conductor polarizable continuum model (CPCM). The reactivity of 14 for insertion of MeNCS and protonation was considered by DFT (Fig. 2). Insertion starts with a simple associative substitution reaction in which one of the coordinating O atoms of the bidentate acetate ligand is replaced by a sulfur atom of the isothiocyanate, TS_14–15, to give κ¹-acetate complex 15 followed by the insertion of MeNCS into the Pd–Ar bond to afford 16. An alternative orientation of MeNCS in TS_15–16′ is higher in energy. For protonation, 14 is substituted by acetic acid to form κ¹-acetate intermediate 17. Proton transfer forms 18. The transition structures for insertion and protonation steps (TS_15–16 and TS_17–18, respectively) both lie lower in energy than those for substitution. TS_14–15 is more stable than TS_14–17 by 1.1 kcal mol⁻¹, suggesting that the insertion reaction would be faster than protonation, a result substantiated by the NMR studies. DFT calculations revealed how the nature of the R groups of RNCS affects the ease of the insertion process (Scheme S3, ESI†). The TS_15–16 is always more stable than TS_15–16′ and the energy barrier dependence of the RNCS substitution (TS_14–15) on the R group is insignificant. The insertion transition structure (TS_15–16) is lower in energy than TS_14–15, with the exception of R = tBu, so the chemoselectivity is dictated by the substitution. For R = tBu, the steric repulsion between the tBu substituent and the aryl ligand in TS_15–16 results in this transition structure being destabilised considerably and lies above TS_14–15. Thus, for R = tBu, the insertion step, and not substitution, is likely to play the crucial role in determining the chemoselectivity. Thus the insertion and protonation pathways become more competitive, which explains the slow reaction of 14 with tBuNCS observed experimentally.

Fig. 2 Energy profile showing competition between protonation and insertion of MeNCS for [(DMSO)Pd(O₂CCH₃)(Ar)], 14. Relative Gibbs and enthalpy energies (in parentheses) are given in kcal mol⁻¹, calculated at the B3LYP-D3BJ/BS3//M06/BS1 level of theory in DMSO using the CPCM approach.

The reaction scope was investigated with different RNCS substrates (Table 1). The resulting N-aryl(alkyl)-2,6-dimethoxythiobenzamide products were characterised by HRMS, ¹H and ¹³C NMR spectroscopy (Fig. S8 and S25, ESI†) and, where possible X-ray crystallography (Fig. S26 and S40, ESI†). The acetic acid formed in eqn (2) can promote protonation (eqn (6)) or other side-reactions so for the reactions with aryl isothiocyanates, K₂CO₃ was added after decarboxylation. The faster insertion of alkyl isothiocyanates meant that addition of base was not required. Following addition of NaBH₄ products 6a–c could be isolated in modest yields (47–57%, Table 1, entries 1–3). The reaction with sterically demanding tBuNCS required a longer reaction time, resulting in more protodecarboxylated and amide products as compared to thioamide 6d. Addition of K₂CO₃ after decarboxylation is essential for reactions with arylisothiocyanates to prevent partial conversion of the thioamides to amides (identified by ESI-MS and isolated and characterised by ¹H NMR for 2,6-dimethoxy-N-phenylbenzamide, Fig. S13 and S22, ESI†). While the exact mechanism of this transformation is unknown, soft metal ions such as Pd²⁺ promote hydrolysis of thioamides to amides¹⁸ and amides have been noted as unwanted side products in palladium catalysed transformations of thioamides.¹⁹ In the case of PhNCS, the yield of 6e improved with the use of both K₂CO₃ and NaBH₄ (Table 1). The yield of isolated thioamide was found to depend on the “H” source added. Addition of 5 equiv. of NaBH₄ gave the optimum yield of 6f and 6g. The potential reduction of the nitro functional group in 6h was avoided by not including a “H” source.

Table 1 Isolated yields for the synthesis of thioamides 6a–h

Entry	Product, R	NaBH4 (equiv.)	Yield (%)
a Reaction in the absence of K2CO3.
1	6a, Me	NaBH4 (5)^a	47
2	6b, Et	NaBH4 (5)^a	50
3	6c, iPr	NaBH4 (5)^a	57
4	6d, tBu	NaBH4 (5)^a	12
5	6e, Ph	NaBH4 (5)	60
6	6f, pClPh	NaBH4 (5)	53
7	6g, pCF3Ph	NaBH4 (5)	40
8	6h, pNO2Ph	None	22

---

In summary, a mechanism-based approach using gas-phase ion chemistry was used to uncover a new transformation of aromatic carboxylic acids to substituted thioamides. This same transformation can be readily achieved as a “one-pot” method in solution using stoichiometric amounts of organic precursors and palladium acetate. Preliminary studies suggest that a range of other aryl carboxylates can be used as substrates.²⁰ The two crucial steps of decarboxylation of the coordinated carboxylate in [(phen)Pd(O₂CC₆H₅)]⁺ (eqn (3)) and insertion of the isothiocyanate into the Pd–C bond of [(phen)Pd(C₆H₅)]⁺ (eqn (4)) are directly related to each other by the isoelectronic nature of CO₂ and RNCS. MacMillan has recently classified decarboxylation reactions in which two fragments recombine as CO₂ExR (ExR = Extrusion–Recombination).²¹ By analogy, the chemistry described here is part of group of CO₂ExIn (ExIn = Extrusion–Insertion) reactions.²² It is possible that this approach is compatible with other small molecules that are isoelectronic with CO₂ and this is currently under investigation.

We acknowledge: Financial support from Australian Research Council (DP1096134 (GNK), DP120101540 (AJC), DP150101388 (AJC, RAJO), DP160100288 and FT130100204 (PSD), computing time from the University of Tasmania and National Computing Infrastructure, Profs. Marisa Kozlowski, Craig Hutton and Jon Tunge for useful discussions and Mr Yang Yang for experimental assistance.

References

1. R. M. Roberts, Serendipity: Accidental Discoveries, Science, Wiley, 1989.
2. (a) A. McNally, C. K. Prier and D. W. MacMillan, Science, 2011, 334, 1114; (b) K. D. Collins, T. Gensch and F. Glorius, Nat. Chem., 2014, 6, 859.
3. (a) Inventing Reactions, ed. L. J. Gooßen, Springer Berlin Heidelberg, 2013; (b) K. N. Houk and P. H.-Y. Cheong, Nature, 2008, 455, 309–313; (c) Q. N. N. Nguyen and D. J. Tantillo, Chem. – Asian J., 2013, 9, 674–680; (d) Y. Wang and Z. X. Yu, Acc. Chem. Res., 2015, 18, 2288–2296; (e) S. Mitsumori, H. Zhang, P. H.-Y. Cheong, K. N. Houk, F. Tanaka and C. F. Barbas, J. Am. Chem. Soc., 2006, 128, 1040–1041.
4. G.-J. Cheng, X. Zhang, L. W. Chung, L. Xu and Y.-D. Wu, J. Am. Chem. Soc., 2015, 137, 1706.
5. C. Iacobucci, S. Reale and F. De Angelis, Angew. Chem., Int. Ed., 2016, 55, 2980.
6. S. A. Kunzi, J. M. S. Toro, T. den Hartog and P. Chen, Isr. J. Chem., 2016, 56, 53.
7. (a) F. Wang, R. Langley, G. Gulten, L. G. Dover, G. S. Besra, W. R. Jacobs, Jr and J. C. Sacchettini, J. Exp. Med., 2007, 204, 73; (b) M. Baumann, I. R. Baxendale, S. V. Ley and N. Nikbin, Beilstein J. Org. Chem., 2011, 7, 442.
8. (a) E. Schaumann, Synthesis of Thioamides and Thiolactams, in Comprehensive Organic Synthesis II, ed. P. Knochel and G. A. Molander, Elsevier, Amsterdam, Netherlands, 2014, vol. 6, p. 411; (b) B. Kaboudin, V. Yarahmadi, J.-Y. Kato and T. Yokomatsu, RSC Adv., 2013, 3, 6435.
9. The formation of CO₂(g) is likely to be a driving force. DFT calculations estimate the energetics as: ΔG = −15.5 kcal mol⁻¹ to form Z-PhC(S)NHCH₃; ΔG = −13.3 kcal mol⁻¹ Z-PhC(S)NHPh. See ESI† Scheme S1.
10. R. A. J. O'Hair and N. J. Rijs, Acc. Chem. Res., 2015, 48, 329.
11. (a) J. Cornella and I. Larrosa, Synthesis, 2012, 653; (b) L. J. Gooßen and K. Gooßen, Top. Organomet. Chem., 2013, 44, 121–142.
12. D. Tanaka, S. P. Romeril and A. G. Myers, J. Am. Chem. Soc., 2005, 127, 10323.
13. J. S. Dickstein, J. M. Curto, O. Gutierrez, C. A. Mulrooney and M. C. Kozlowski, J. Org. Chem., 2013, 78, 4744.
14. J. Louie, Curr. Org. Chem., 2005, 9, 605.
15. J. Vicente, J.-A. Abad, R. Bergs, M. C. Ramirez de Arellano, E. Martínez-Viviente and P. G. Jones, Organometallics, 2000, 19, 5597.
16. (a) M. A. Ortuño, S. Conejero and A. Lledós, Beilstein J. Org. Chem., 2013, 9, 1352; (b) M. Woolley, A. Ariafard, G. N. Khairallah, K. H.-J. Kwan, P. S. Donnelly, J. M. White, A. J. Canty, B. F. Yates and R. A. J. O'Hair, J. Org. Chem., 2014, 79, 12056.
17. G. Sipos, E. E. Drinkel and R. Dorta, Chem. Soc. Rev., 2015, 44, 3834.
18. D. P. N. Satchell and R. S. Satchell, in The Chemistry of Sulphur-Containing Functional Groups, ed. S. Patai, Z. Rappoport, Wiley, Chichester, 1993, ch 12, p. 599.
19. (a) K. Inamoto, C. Hasegawa, K. Hiroya and T. Doi, Org. Lett., 2008, 10, 5147; (b) T. Yamauchi, F. Shibahara and T. Murai, Org. Lett., 2015, 17, 5392.
20. Substituted coordinated benzoates [(phen)Pd(O₂CC₆H₄X)]⁺ (e.g. X = o-OMe, p-OMe and o-NO₂) also undergo decarboxylation followed by insertion in both the gas and condensed phase consistent with N. Rameau, S. Cadot, A. Paquet, C. Pinel and L. Djakovitch, Top. Catal., 2014, 57, 1430 ).
21. C. C. Le and D. W. C. MacMillan, J. Am. Chem. Soc., 2015, 137, 11938.
22. For “intercepted decarboxylative allylation” see: J. D. Weaver, A. Recio, A. J. Grenning and J. A. Tunge, Chem. Rev., 2011, 111, 1846.

---

Footnote

† Electronic supplementary information (ESI) available. CCDC 1468845–1468850. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c7cc00865a

---