An alkyne trimerisation reaction is a 2+2+2 cyclization reaction in which three alkyne molecules react to form an aromatic compound. The reaction is 'pseudo' pericyclic since it has not been observed to occur without the assistance of metal catalysis; and the metal catalyst assembles the ring stepwise via intermediates which are not directly in between (in a geometric sense) the starting material and products.[1]
Contents |
In 1866, Berthelot reported the first example of cyclotrimerization leading to aromatic products, the cyclization of acetylene to benzene.[2] The reaction required very high temperatures to proceed (400 °C) and produced a complex mixture of products, and research in the field remained dormant until the late 1940s. In 1948, Reppe discovered that nickel could catalyze the formation of substituted benzenes from acetylenic compounds;[3] since his initial discovery, cyclotrimerizations to produce substituted benzenes have been catalyzed by no less than seventeen transition metals, including:
Advantages: Control of the substitution pattern of the aromatic product is good in many cases, and cyclotrimerization can be used in cases when functionalization of pre-formed aromatic materials (through electrophilic aromatic substitution, for instance) is not selective. The reaction is highly chemoselective for triple bonds and can tolerate a wide variety of functional groups on the starting materials.
Disadvantages: For cotrimerizations involving two or three different acetylenes, a variety of regioisomers may form. Operationally, these reactions usually require elevated temperatures (>60 °C) and sometimes require irradiation to facilitate the dissociation of strongly binding carbon monoxide ligands. Catalyst deactivation can occur through the formation of stable, 18-electron η4-complexes incorporating cyclobutadiene,[11] cyclohexadiene, and arene ligands. The most problematic side products of the reaction are due to cyclotetramerization (leading to cyclooctatetraenes) and alkyne dimerization (leading to enynes).
The most common mechanism for the cyclotrimerization of acetylenes begins with the formation of a metallocyclopentadiene complex. Oxidative cyclization of two coordinated alkyne units produces either metallocycle 3 or 4.[12] After dissociation of a third ligand and coordination of a third alkyne, two pathways are possible. Either alkyne insertion generates metallocycloheptatriene 5, or [4+2] cycloaddition generates bridged bicycle 6. The former pathway is questionable however, as reductive elimination from a metallocycloheptatriene to form an arene is symmetry forbidden.[13] A third pathway proposed for ruthenium catalysts involves formal [2+2] cycloaddition of the alkyne followed by rearrangement, reductive elimination, and arene decomplexation. The intermediacy of bicycle 7 was supported by DFT calculations.[14]
For cyclotrimerization reactions of asymmetrically substituted acetylenes, a number of regioisomeric products are possible. The substitution pattern about the product arene depends on two events: formation of the metallocyclopentadiene intermediate and incorporation of the third equivalent of alkyne. Although both head-to-head and tail-to-tail metallocyclopentadienes lead to 1, a number of acetylenic substrates selectively form regioisomers of type 2. Steric bulk on the alkyne coupling partners and catalyst have been invoked as the controlling elements of regioselectivity.[15]
Chiral catalysts have been employed in combination with arynes to produce non-racemic amounts of atropisomeric products.[16] Selective cyclotrimerization of one of a pair of enantiotopic alkynes has also been facilitated by a chiral catalyst.[17]
Catalysts for cyclotrimerization are highly selective for triple bonds, which gives the reaction a fairly wide substrate scope. Alcohols, ethers, amines, and carbonyl compounds (ketones, esters, amides, and carboxylic acids) are well tolerated. Other than carbonyl compounds, the reaction is not generally compatible with unsaturated functionality. Nitriles, for instance, can react to form pyridines.
The reaction is limited in some cases by catalyst deactivation due to the formation of stable, 18-electron η4-complexes.[18] Cyclobutadiene, cyclohexadiene, and arene complexes have all been observed as off-cycle, inactivate catalyst forms. In addition to high-order polymers and dimers and trimers, which originate from low regio- and chemoselectivities, enyne side products derived from alkyne dimerization have been observed. Rhodium catalysts are particularly adept at enyne formation (see below).[19] For nickel catalysis, formation of larger rings (particularly cyclooctatetraene) can be a problem.
Cyclization of diynes with a separate alkyne can provide fused ring systems with high atom economy. In a striking example in the synthesis of estrone, diyne 8 was combined with di(trimethylsilyl)acetylene to produce benzocyclobutane 9.[20] Upon heating, ring-opening produced a quinone methide that participated in a subsequent intramolecular Diels-Alder reaction.
If the third alkyne unit is tethered to the first two, three rings can be created in a single step. In the example below in the synthesis of calomelanolactone, Wilkinson's catalyst was used to catalyze an intramolecular cotrimerization of triyne 10.[21]
Crowded triynes can cyclize to products exhibiting helical chirality. In one example, highly favorable due to the formation of three new aromatic rings in one step, triyne 12 transforms to helical product 13 in the presence of cyclopentadienylcobalt dicarbonyl.[22] As of 2004, this process had yet to be rendered asymmetric, but the products could be separated through chiral HPLC.
Cyclotrimerization presents an alternative to the functionalization of pre-formed aromatic compounds through electrophilic or nucleophilic substitution, the regioselectivity of which can sometimes be difficult to control.
Other methods for the direct formation of aromatic rings from substituted, unsaturated precursors include the Dötz reaction, palladium-catalyzed [4+2] benzannulation of enynes with alkynes,[23] and Lewis-acid-mediated [4+2] cycloaddition of enynes with alkynes.[24] Cyclization of transient benzyne species with alkynes, catalyzed by palladium, can also produce substituted aromatic compounds.[25]
Although various experimental conditions have been used to perform this reaction, commercially available CpCo(CO)2 remains the most common catalyst used. High temperatures and irradiation with light are often required in order to facilitate the dissociation of carbon monoxide from the catalyst.
1,2-Bis(propiolyl)benzene (350 mg, 2.3 mmol) was dissolved in diglyme (8 mL) and added over a period of 36 hours (with a syringe pump) to a refluxing solution of bis(trimethylsilyl)acetylene (5 mL) and CpCo(CO)2 (60 µL) in n-octane (70 mL), under a dry N2 atmosphere. Removal of the solvent and unreacted BTMSA by vacuum transfer followed by column chromatography on silica gel (100 g) (pentane:ether = 80:20) gave a yellow oil which solidified on standing. Recrystallization of the crude solid from methanol gave orange crystals of the title anthraquinone (120 mg, 15 %), mp 69–71.5°: IR (CCl4) 2900, 1670, 1540, 1320, 1290, 1250, 1005, 980, 952, 932 cm–1; 1H NMR δ 1.50 (s, 2H), 1.70 (m, 2H), 2.23 (m, 2H), 9.51 (s, 18H); EIMS m/z (relative intensity): M+ 352 (14), 337 (17), 73 (67), 57 (79), 55 (77), 43 (100), 41 (67).