[15489-27-7] · Cl4CuLi2 · Dilithium Tetrachlorocuprate(II) · (MW 219.24)
(organic soluble copper catalyst1 used in catalytic quantities to improve the efficiency of halide2 and allylic acetate3 displacements with Grignard reagents; it is used in stoichiometric amounts to open epoxides4 and aziridines5a)
Physical Data: d 0.910 g cm-3.
Solubility: sol up to concentrations of 0.5 M in THF.
Form Supplied in: commercially available as a 0.1 M solution in THF.2a
Preparative Methods: prepared in THF solution by reaction of Lithium Chloride (0.2 mol) and Copper(II) Chloride (0.1 mol).5b
Handling, Storage, and Precautions: moisture and oxygen sensitive.
The addition of two or more equivalents of lithium chloride to copper(II) chloride will result in a dark red THF soluble solution of dilithium tetrachlorocuprate.1,2 Dilithium tetrachlorocuprate has been used most frequently to enhance the efficiency of halide displacements with a Grignard reagent. The reagent is used catalytically in amounts less than 10 mol % and typically it will increase the displacement yield.
The first report describing the use of dilithium tetrachlorocuprate was by Kochi and Tamura in 1971.2 They described the coupling of n-butylmagnesium bromide with n-hexyl bromide using Li2CuCl4 to afford decane in 78% yield. This report was followed by the use of more complex substrates, including Grignard reagents containing protected ethers, unsaturation, and strained rings with homoallylic halides (eqs 1-3).6-8 It is noteworthy that displacement predominated over dehydrohalogenation in these cases.
Sterically hindered Grignard reagents were coupled with toluenesulfonates to afford sterols with modified side chains (eq 4)9 and with benzylic halides resulting in neopentyl derivatives of naphthalenes (eq 5).10 In this example the copper catalyst reduced homodimer formation (i.e. 1,2-bisnaphthylethane formation).
The benzylic type Grignard reagent derived from 3-chloromethylfuran would not displace an allylic halide unless dilithium tetrachlorocuprate was employed (eq 6).11
Functionalized Grignard reagents were efficiently coupled with halides containing an ester (eq 7)12 or carboxylic acid salt (eq 8).13 In both cases there was little reaction at the carboxyl termini. The alkenic product from eq 8 was of interest for use in ultra thin-layer photoresists. High purity alkene in which double bond isomers were not present was obtained with Li2CuCl4 catalysis.
A vinylic type Grignard reagent was coupled with n-octyl iodide to afford 2-substituted 1,3-dienes in high yields. In this example, CuI was not an effective coupling catalyst (eq 9).14
A double displacement was reported by Schlosser using 1 mol % of the copper catalyst (eq 10).15
A cyclic a-halo ether was displaced with the aid of dilithium tetrachlorocuprate. The copper catalyst increased the yield and stereoselectivity for the displacement (eq 11).16
The first report using dilithium tetrachlorocuprate in allylic acetate displacements was by Schlosser in which trans-crotyl acetate was converted into trans-2-octene (eq 12).3 The reaction was reported to be regio- and stereoselective. In addition to the desired coupling product, the reaction also produced 10% of the carbonyl addition product 5-methyl-5-nonanol. The level of carbonyl addition product could be reduced using stoichiometric amounts of Copper(I) Iodide. The coupling of a (Z)-trisubstituted allylic acetate also proceeded with 100% regio- and stereocontrol (eq 13).17 In this latter example, nerol derivatives containing an allylic acetate were compared with those containing a bromide, sulfonium salt, ammonium salt, and a carbonate. The acetate was superior, both in terms of maintaining alkene geometry and regiocontrol (a- vs. g-attack).
The regiocontrol was rationalized by Backvall, who observed drastic changes in regiochemistry with changes in reaction conditions. Rapid addition of the Grignard reagent resulted in a-attack (eq 14).18 A slow Grignard addition in the presence of higher catalyst loadings (5 mol % vs. 2 mol %) completely reversed the regiochemistry of the displacement.
These results were explained by assuming the presence of a monoalkylcopper complex (1) and a dialkylcopper complex (2) (eq 15). With rapid addition, there is a buildup of the dialkylcopper species which favors the a-product. Support for this was provided by reacting preformed n-Bu2CuLi with the acetate to afford a 96:4 mixture of regioisomers. A slow addition, on the other hand, will result in predominant reaction via the monoalkylcopper species, which favors g-attack.
The displacement of allylic acetates was extended to diene systems. A study of all four stereoisomers of 2,4-heptadienyl acetate showed that only the (E,E) isomer was 100% stereospecific (eq 16).19 The (2Z,4E) isomer was the least stereospecific, resulting in a 2:1 ratio of diene isomers (eq 17).
The displacement reported by Roush in which an (E,E)-diene afforded a mixture of regioisomers may have been improved using the rapid addition technique or by using a preformed copper reagent (eq 18).20
An allylic quaternary ammonium salt was also a good substrate for displacement under dilithium tetrachlorocuprate catalysis (eq 19).21
When used in excess of stoichiometric quantities, the reagent will open epoxides4 and activated aziridines.5a The epoxide of cyclohexene was opened at rt to afford a trans-chlorohydrin in near quantitative yield (eq 20).4 The opening usually occurs at the least hindered site; however, styrene oxide produced a mixture of chlorohydrins with the secondary chloride predominating. The opening of strained rings was extended to aziridines which are activated as the toluenesulfonamide (eq 21).5
There are only a few reports describing the use of dilithium tetrachlorocuprate as a catalyst for coupling Grignard reagents with vinylic and aryl iodides.14,22-24 In one report, one of two iodides was replaced using Li2CuCl4 (eq 22).23 An example using a 1,3-diene indicates that palladium catalysis may be more effective for these types of transformations (eq 23).14
There is one report in which a carbonylative cyclization was observed, presumably by intramolecular trapping of an acylmetal derivative (eq 24).24
Andrew S. Thompson
Merck & Co., Rahway, NJ, USA