Synthesis of α-diketones from alkylaryl- and diarylalkynes using mercuric salts.

Both alkylarylalkynes and diarylalkynes 1 are converted into the α-diketones 2 in good yield by the use of mercuric salts, e.g., mercuric nitrate hydrate or mercuric triflate, in the presence of water. Other mercuric salts, e.g., sulfate, chloride, acetate, or trifluoroacetate, do not provide the diketone product. A possible mechanism is proposed.

The oxidation of alkylarylalkynes and diarylalkynes 1 to furnish the corresponding α-diketones 2 is well-known in organic chemistry. A very large number of oxidants have been used for this process. For example, KMnO4 has been employed often for this transformation[1] as has RuO4 (often generated in situ or immobilized).[2] There are also several reports of the use of various DMSO-based oxidations, usually with an added electrophile[3] or in the presence of a palladium catalyst[4] for the formation of 2 from 1. Finally, a large variety of other metal-based[5] and nonmetal-based[6] oxidations have been reported. For a project involving the synthesis of androgen receptor antagonists, we had need of a good method for converting alkylarylalkynes into α-diketones. We report here that methodology and its application to the conversion of several disubstituted alkynes 1 to the corresponding α-diketones 2.

We hoped if it might be possible to intercept the well-known mechanism[7] for mercuric-catalyzed hydration of an alkyne (Scheme 1) by reaction of the α-mercurioketone intermediate D with another equivalent of the mercuric salt.

Scheme 1

Mercuric Nitrate Catalyzed Hydration of an Alkyne

We examined this process using 1,2-diphenylacetylene 1a as the substrate. Thus treatment of 1a with 2 equiv of mercuric nitrate hydrate at 22 °C in aqueous THF for 12 h gave the expected α-diketone, benzil 2a, in 52% yield (Scheme 2). We carried out several experiments on this test reaction to find the best set of conditions. Carrying out the reaction under air or under argon gave the same results, so oxygen is not required for the process. The use of anhydrous THF with an added equivalent of water afforded good yields. Other solvents worked well, e.g., methanol, DME, dioxane, acetonitrile, acetone, acetic acid, DMSO, and especially DMF. The best yields were obtained in methanol with 1 equiv of added water (22 °C, 20 min, 84%) and in DMF (22 °C, 24 h, 90%). The use of other mercuric salts, e.g., sulfate, chloride, acetate, and trifluoroacetate, did not give any 2a. However, the use of mercuric triflate, Hg(OTf)2, also produced good yields of the α-diketone 2a.

Scheme 2

Oxidation of 1a To Give 2a

We then applied this method to the synthesis of a wide variety of α-diketones using the following set of conditions, namely treatment of the alkyne 1 with 2 equiv of mercuric nitrate hydrate in DMF in air at 22 °C. The results are shown in Table 1. Arylalkynes generally gave good yields of the expected α-diketone products, e.g., 2b,d–f,m,n, with yields ranging from 47 to 82%. In addition, most of the diarylacetylenes gave quite good yields of product. The presence of halogens (1c,i,j,n) did not hinder the oxidation nor did the more oxidizable or labile functionalities, such as phenols (1l and 1m), esters (1k and 1o), amines (1e, 1p, and 1q), or nitriles (1f), all of which gave reasonable yields of the desired products 2. Substrates with electron-donating substituents (methoxy, 2g, or methyl, 2h) gave good yields, while the substrate having a 4-trifluoromethyl group 1d gave the lowest yield among the successful substrates, perhaps due to lower electron density in the alkyne. The 1,4-di(propynyl)benzene 1r afforded the bis(α-diketone) 2r in good yield. A few substrates did not work well in this reaction, giving mixtures of products. Thus, the 2-nitrophenyl substrate 1s gave the expected α-diketone 2s in 33% yield along with the expected 3-benzoylanthranil 3s in 54% yield. The analogous 2-nitrophenyl substrate with a propyl group on the end of the alkyne, 1t, gave none of the α-diketone and only the anthranil 3t in 61% yield. This cyclization of a 1-(2-nitrophenyl)alkynes such as 1s and 1t is well-known[8] and is usually carried out by treatment with either transition metals or Lewis acids. We also attempted the oxidation on two different heterocyclic alkyne substrates. Thus, the protected 4-pentynyl imidazole 1u gave a relatively good yield of the desired α-diketone 2u (69%) accompanied by the simple hydration product 3u in 20% yield. The 3-pentynylpyridine 1v likewise gave the desired α-diketone 2v as the major product (49%) along with the opposite hydration product 3v in 38% yield.[9] Finally the symmetrical (2-nitrophenyl)acetylene 1w did not give any of the desired α-diketone product 2w but rather the two byproducts, the simple hydration product 3w in 32% yield and the anthranil 3w′ in 37% yield.[10]

Table 1

Oxidation of Alkynes with Mercuric Nitrate Hydrate

We believe that the mechanism involves the steps shown in Scheme 3, namely attack of the alkyne 1 on the mercuric nitrate to give the cyclic mercuronium ion A, which is then attacked by water to give B, which loses a proton to give the α-mercurio enol C. Tautomerization would then give the α-mercurio ketone D. Up to this point, this is the same mechanism as for the simple hydration of the alkyne (as shown in Scheme 1). The key step is the attack of nitrate on the α-mercurio ketone D, with activation by the second equivalent of mercuric nitrate, to generate the α-nitrato ketone H and mercurous nitrate. The final step is the reductive elimination of nitrous acid from H to give the observed α-diketone product 2.

Scheme 3

Proposed Mechanism of Oxidation with Mercuric Nitrate

Perhaps the most unusual step in this proposed mechanism is the conversion of the α-mercurio ketone D to the α-nitrato ketone H, but this step has precedent in the literature since a similar conversion of 2-methoxy-1,2-diphenylethyl mercuric nitrate to the 2-methoxy-1,2-diphenylethyl nitrate is known.[11] There is also good precedent for the final step, since the conversion of α-nitrato ketones to α-diketones is well-known.[12]

This oxidation also proceeds, although less well, with 2 equiv of anhydrous mercuric triflate and 2 equiv of water in THF. We propose a very similar mechanism for the formation of the α-diketone 2 in this reaction (Scheme 4), namely the attack of triflate ion on the corresponding α-mercurio ketone D′ activated by mercuric triflate to give the α-sulfonyloxy ketone H′. Elimination of trifluoromethanesulfinate from H′ would then give the α-diketone 2. This last step, the elimination of sulfinates to give ketones, is well precedented in the literature.[13]

Scheme 4

Proposed Mechanism of Oxidation with Mercuric Triflate

In summary, we have developed a new method for the oxidation of alkylarylalkynes and diarylalkynes 1 to give α-diketones 2 with mercuric salts. The reaction is limited to salts that can undergo facile subsequent elimination, namely nitrates and triflates. The use of this process in synthesis is underway and will be reported in due course.