Unprecedented carbon dioxide effect on a Pd-catalysed oxidative carbonylation reaction: a new synthesis of pyrrole-2-acetic esters Bartolo Gabriele,* a Giuseppe Salerno,* b Alessia Fazio b and Fausto Bruno Campana b a Dipartimento di Scienze Farmaceutiche, Università della Calabria, 87036 Arcavacata di Rende (CS), Italy. E-mail: b.gabriele@unical.it; Fax: +39 0984 492044; Tel: +39 0984 492130 b Dipartimento di Chimica, Università della Calabria, 87036 Arcavacata di Rende (CS), Italy. E-mail: g.salerno@unical.it; Fax: 39 0984 492044; Tel: +39 0984 492066 Received (in Cambridge, UK) 9th April 2002, Accepted 14th May 2002 First published as an Advance Article on the web 31st May 2002 It has been found that carbon dioxide effectively promotes the Pd-catalysed oxidative cyclisation–alkoxycarbonylation of (Z)-(2-en-4-ynyl)amines 1 leading to pyrrol-2-acetic esters 2. We recently described a new synthesis of furan-2-acetic esters by PdI 2 /KI-catalysed oxidative cyclisation–alkoxycarbonyla- tion of (Z)-2-en-4-yn-1-ols. 1 When we applied this method- ology to (Z)-(2-en-4-ynyl)amines 1, however, the desired pyrrole-2-acetic esters 2 were consistently obtained in low yield, even after optimisation of the reaction conditions. For example, carbonylation of (Z)-butyl(2-ethylnon-2-en-4-ynyl)- amine 1a under 40 atm of a 3+1 mixture of carbon monoxide and air in MeOH at 70 °C for 5 h (substrate/KI/PdI 2 molar ratio = 100+200+1, substrate concentration = 0.05 mmol mL 21 MeOH) afforded a mixture of methyl 2-(1-butyl-4-ethyl-1H- pyrrol-2-yl)hexanoate 2a (45% GLC yield, 40% isolated) and 1-butyl-4-ethyl-2-pentyl-1H-pyrrole 3a (deriving from a com- petitive cycloisomerization process, 2 9% GLC yield) at total substrate conversion [eqn. (1)]. (1) The striking difference in the behaviour of enynols and enynamines can be ascribed to the basicity of the latter. 3 The alkoxycarbonylation process occurs with reduction of PdI 2 to Pd(0) and formation of 2 mol of HI. Pd(0) is then reoxidised to the catalytically active species PdI 2 by I 2 , formed by oxidation of HI with oxygen 4 (Scheme 1; anionic iodide ligands are omitted for simplicity). The latter process is very fast in the case of (Z)-2-en-4-yn-1-ols or in general with non-basic substrates. However, substrates that are basic enough to be protonated by HI, such as enynamines, can efficiently inhibit the reoxidation of Pd(0) and therefore hinder the overall oxidative carbonyla- tion process. When this occurs, the substrate may undergo undesired side-reactions such as oligomerisation and/or oxida- tive degradation. One way for accelerating the Pd(0) reoxidation process when inhibited by a basic substrate consists of the use of a large excess of oxygen (which apparently favours the oxidation of HI to iodine). 3b A large excess of oxygen for the carbonylation of enynamines, however, could not be used owing to the tendency of the pyrrole ring to undergo oxidation reactions. 5 It was therefore necessary to solve the problem in a different manner. A reactant able to reversibly bind the amino group [thus ‘freeing’ the HI necessary for the reoxidation of Pd(0)] without hampering the cyclisation–alkoxycarbonylation process was needed. We have found that carbon dioxide effectively fulfils these requirements, through the formation of a carbamate species. The nitrogen in the carbamate, while much less basic than in the substrate, may still act as nucleophile, since CO 2 can be eliminated during the cyclisation process (Scheme 2). In fact, by reacting 1a under the above-mentioned conditions but with the addition of 50 atm of CO 2 , 2a was obtained in 70% GLC yield (62% isolated) at total substrate conversion, without any formation of 3a [eqn. (2)].† In order to support our hypothesis regarding the effect exerted by CO 2 , we synthesized allyl carbamate 4 and let it react under the above-mentioned conditions but in the absence of CO 2 . Palladium-promoted cleavage of the allyl group of 4 was expected to afford exactly the same carbamate intermediate formed in the CO 2 -assisted carbonylation of 1a (Scheme 3).‡ Indeed, by reacting 4 in the absence of CO 2 and under 6 atm of a 5+1 mixture of CO and air, after 48 h 2a was obtained in 46% GLC yield at total substrate conversion, with complete absence of pyrrole 3a.§ In agreement with the mechanism shown in Scheme 3, the allyl moiety of 4 was recovered as methyl but-3-enoate. Under the optimised conditions found for 1a, other (Z)-(2-en- 4-ynyl)amines 1 were efficiently carbonylated to selectively afford the corresponding pyrrole-2-acetic esters 2 in good yields. For example, methyl (1-butyl-4-ethyl-1H-pyrrol-2-yl)- acetate 2b was obtained in 72% GLC yield (65% isolated) from the reaction of (Z)-butyl[2-ethyl-5-(trimethylsilanyl)pent-2-en- 4-ynyl]amine 1b at total substrate conversion [eqn. (2)].† (2) Scheme 1 Mechanism of the PdI 2 /KI-catalysed oxidative carbonylation of 1a to give 2a. Anionic iodide ligands are omitted for clarity. This journal is © The Royal Society of Chemistry 2002 1408 CHEM. COMMUN. , 2002, 1408–1409 DOI: 10.1039/b203413a