The addition of terminal alkynes to aldehydes and ketones, especially in an enantioselective manner, is of great interest because of the versatility of the corresponding propargylic alcohols. Stoichiometric amounts of strong bases such as organolithium or dialkylzinc reagents are widely used for this type of reaction with or without chiral ligands or chiral Lewis acid complexes. Intrinsic drawbacks, however, such as the use of stoichiometric amounts of metal reagents and a separate metal acetylide preparation step, make it difficult to achieve an atom-economical process.

Given the recent strong demand for an environmentally-benign process with high total efficiency, the in situ catalytic generation of metal nucleophiles and their use in carbon-carbon bond-forming reactions is currently a major interest in organic synthesis. The use of only catalytic amounts of chiral metal salts to achieve truly catalytic asymmetric reactions using terminal alkynes directly as a substrate is eagerly anticipated. Carreira and coworkers reported the sophisticated example of a catalytic system of Zn（OTf）2, N-methylephedrine, and Et3N.1 Due to the Cannizzaro reaction, however, aromatic aldehydes cannot be used in this catalytic system. Another catalytic alkynylation of aldehydes and ketones using a catalytic amount of strong hydroxide, alkoxide, or phosphazene base in polar solvents has also been reported, although the substrate generality is still limited and there has been no application to the catalytic enantioselective reactions using these bases. Thus, there remains much room to develop catalytic alkynylation of various carbonyl compounds under mild conditions.

Development of a New catalytic System for the Alkynylation of Aldehydes and Ketones2

Shibasaki's group has developed various bifunctional catalysts, such as heterobimetallic catalysts and Lewis acid-Lewis base catalysts, to achieve efficient enantioselective reactions under mild conditions with minimal undesired waste. Application of this bifunctional strategy seems to be one of the most promising solutions for developing a catalytic alkynylation of a broad range of aldehydes and ketones. From this point of view, dual activation of soft pro-nucleophiles （terminal alkynes）and hard electrophiles （carbonyl compounds）is very important. Indium（III） salts are efficient Lewis acids for carbonyl compounds. Quite recently, indium（III）salts have emerged as effective activators of alkynyl groups in cross-coupling reactions, etc. This “bifunctional character” of indium（III）prompted me to examine indium（III）salts for the alkynylation of carbonyl compounds via dual activation （Scheme 1）.

Using cyclohexanecarboxaldehyde （1a） or benzaldehyde （1b） with phenylacetylene （2a）as representative substrates, I started to examine alkynylation of aldehydes using indium（III） salts under various conditions and found that the combination of InBr3 with i-Pr2NEt provided the optimal reaction efficiency in the alkynylation of aldehydes. As summarized in Table 1, the optimized reaction conditions were applicable for a wide range of aldehydes.

The use of ketones instead of aldehydes, however, gave the corresponding tertiary propargylic alcohols in very low yield under these conditions. For ketones, In（OTf）3 （20 mol %） was found to be the indium source of choice and the results are shown in Table 2. Just changing the indium source, these catalytic systems provided the broad substrate scope, including ketones, under mild conditions.2

Dual activation of both soft pro-nucleophiles and hard electrophiles is the key to this reaction and was confirmed in the mechanistic studies using in situ IR and NMR spectroscopic analysis. First, to gain information about the activation of alkyne, in situ IR spectra were measured. When 1 equiv of phenylacetylene was added to the solution of InBr3 and i-Pr2NEt, there was a signal at 3246 cm-1 corresponding to the C-H stretch of the alkyne, and this signal disappeared in less than 1 min. On the other hand, when 1 to 3 equiv of 2a was added （total 2 to 4 equiv）, the absorbance at 3246 cm-1 increased with each addition. These results suggested that InBr3 activated the terminal alkyne and that the indium monoacetylide species was formed. Next, to confirm the activation of the carbonyl compound by Lewis acidic indium（III） salt, NMR spectroscopic analysis was performed using InBr3. The shift of the peak corresponding to the aldehyde proton （in 1HNMR spectrum） and carbonyl carbon （in 13CNMR spectrum） was observed following the addition of InBr3 to aldehyde in the presence or absence of i-Pr2NEt and phenylacetylene （2a）, indicating the activation of the carbonyl compound by coordination to the indium（III）species.

Development of Catalytic Asymmetric Alkynylation of Aldehydes3

The success in developing mild reaction conditions prompted me to further develop asymmetric variants to produce versatile optically-active propargylic alcohols. Initial studies on the development of the asymmetric reaction conditions revealed that the use of BINOL as a chiral ligand had high enantioselectivity in the addition of phenylacetylene（2a）to cyclohexanecarboxaldehyde（1a）;in the presence of 10 mol % InBr3, 10 mol %（R）-BINOL（1:1 ratio）, and 50 mol % i-Pr2NEt in CH2Cl2 at 40℃, the propargylic alcohol 3aa was obtained in 96% ee, although the chemical yield was moderate （46%, after 7 h）. Further optimization of reaction conditions led to the finding that the use of Cy2NMe instead of i-Pr2NEt effectively accelerated the reaction, giving the product in 84% yield and 98% ee （after 7 h）.

The generality of this catalytic system （10 mol % InBr3 and （R）-BINOL, and 50 mol % Cy2NMe in CH2Cl2 at 40℃）was examined using aliphatic aldehydes, as summarized in Table 3. Even using the less reactive alkylacetylene 2b, 2d, and 2e instead of phenylacetylene （2a）, good chemical yield was obtained with excellent enantioselectivity （entries 2-4）. Although aldehydes with a primary alkyl group could not be utilized in racemic system, the reaction with isovaleraldehyde （1l） also proceeded under the same conditions in highly enantioselective manner （entries 5,6）. Even for the very easily enolizable aldhehyde, hydrocinnamaldehyde （1m）, slow addition of the aldehyde prevented side reactions such as self-condensation, providing the desired product in good yield and excellent enantioselectivity （entry 7）.

Furthermore, the optimized conditions were also applicable for aromatic aldehydes, which are quite challenging substrates for existing catalytic systems due to a competitive Cannizzaro reaction （Table 4）. The addition of phenylacetylene （2a） to benzaldehyde （1b） proceeded smoothly to give the corresponding product 3ba in 84% yield and 95% ee after 24 h. The use of the alkyl- and alkenylacetylenes also produced high enantioselectivity （entries 2-5）. Especially, the accommodation of alkylalkynes having some functional group, such as 2d, is noteworthy. In addition, benzaldehyde derivatives with the electrondonating substituent or electron-withdrawing substituent gave satisfactory yields and high enantioselectivity（entries 6-8）. Heteroaromatic aldehydes, such as 3-thiophenecarboxaldehyde（1k） or 3-furaldehyde （1n） can also be utilized as electrophiles （entries 9,10）.

Lowering the catalyst loading was possible; the use of 1-2 mol % of catalyst promoted the reaction of benzaldehyde （1b） with 2a, affording good yield and enantioselectivity, while longer reaction time was necessary. Notably, the reaction also proceeded under near conditions in highly enantioselective manner with 1 mol % of catalyst （Scheme 2）.

To gain insight into the reaction mechanism, when the reaction was performed using nonenantiopure BINOL, rather strong positive nonlinear effects were observed between the enantiomeric excess of BINOL and the product （Figure 2）. Several experiments revealed that the origin of nonlinear effects is a selective resolution process. The heterochiral complex derived from InBr3, BINOL, and Cy2NMe selectively caused a precipitaion. Thus, large amplification was observed in the reaction; for example, the use of 20% ee of BINOL afforded the propargylic alcohol 3aa in 94% ee. These experiments also implied the reactivity difference by the structure of amine base; the differenc of amine base would affect on the property or stability of catalyst complex.

Although the precise mechanism is not clear at present, based on the dual activation mechanism that was confirmed in racemic system, proposed catalytic cycle is illustrated in Scheme 3; first, indium acetylide species would be generated via the activation of alkyne by indium complex. It could react with aldehyde that would also be activated by indium/BINOL complex. The resulting indium alkoxide should be protonated, leading to the regeneration of catalyst with releasing the propargylic alcohol. In the addition step, the monometallic mechanism cannot be completely excluded. The bimetallic mechanism, however, would be much more plausible by the following reasons; （1） The results obtained from NLE experiments （cf. Figure 2） as well as related experiments suggested the involvement of the bimetallic mechanism. （2） The reactivity is highly dependent on the ligand structure. （3） Intramolecular addition mechanism in the monometallic mechanism seems to be difficult. More precise mechanistic studies, including X-ray crystarographic analysis of catalyst, are ongoing.

I developed a new catalytic alkynylation of aldehydes and ketones by focusing on “bifunctional character” of indium（III）. Dual activation of both soft nucleophiles and hard electrophiles is the key to this reaction and was confirmed by in situ IR and NMR spectroscopic studies. The asymmetric variant of this reaction was promoted by the In（III）/BINOL complex, giving the propargylic alcohols in highly enantioselective manner （83->99% ee）. The striking feature is that this catalyst system is very simple, mild, and applicable for a wide range of substrates. It provides a new entry in bifunctional catalysis.

Scheme 1. Alkynylation via Dual Activation of Both CarbonylCompounds and Alkyne

Table 1. InBr3-Catalyzed Alkynylation of Aldehydes

Table 2. In（OTf）3-Catalyzed Alkynylation of Ketones

Figure 1. In situ IR study of the successive addition of phenyl-acetylene（2a）to InBr3 and i-Pr2NEt in DME. a）the C-H stretchsignal of 2a, b）time course of the successive addition（1-4 equivof phenylacetylene）at 3246 cm-1.

Table 3. InBr3/BINOL Complex-Catalyzed Asymmetric Alkynylation　of Aliphatic Aldehydes

Table 4. InBr3/BINOL Complex-Catalyzed Asymmetric Alkynylationof Aromatic Aldehydes

Scheme 2. The Reaction Performed with 1 mol % of Catalyst under Neat Conditions

Figure 2. （+）-Nonlinear Effects in Asymmetric Alkynylation Catalyzed by InBr3/BINOL Complex

Scheme 2-9. Proposed Catalytic Cycle

末端アルキンのカルボニル化合物への不斉付加反応は効率的にキラルプロパルギルアルコールを与える有用な炭素一炭素結合形成反応である。その多くは金属試薬を両論量用いて達成されてきたが、効率性・環境調和性の高いプロセスの構築には触媒反応の開発が望ましい。Carreiraらが本反応の初の触媒的不斉化を報告しているものの、基質一般性において改善の余地を残していた。

滝田良は一般性の高い触媒系の構築を目指し、求核剤と求電子剤の両方を活性化させるデュアルアクティベーションというコンセプトを基盤に不斉反応の開発を行った（Scheme 1）。末端アルキンとカルボニル化合物の両方の活性化を達成するためにはソフト性とハード性を持ち合わせた金属触媒の構築が必要である。滝田良はインジウムの"bifunctional character"とも言える性質に着目し、詳細な検討を行った。その結果InBr3、とi-Pr2NEtの組み合わせが、アルデヒドに対する末端アルキンの付加反応に有効な触媒系であることを見いだした（Table 1）。さらにより困難であるケトンに対する付加反応はInBr3、に替えてIn（OTf）3を用いることにより有効な触媒系を構築することに成功した。

想定したデュアルアクティベーション機構はin situ IR及びMRを用いた解析実験により明らかにすることができた。

これらの結果を基に、続いて本反応の不斉化の検討を行った。初期検討においてBINOLを不斉配位子として用いることにより目的のプロパルギルアルコールが高い不斉収率で得られたが、反応性の低さが問題となった。そこでさらなる詳細な検討を行ったところ、塩基としてCy2NMeを用いることにより、高い反応性が得られることがわかった。

開発した触媒系（10 mol % InBr3 and（R）-BINOL, and 50 mol % Cy2NMe in CH2Cl2 at 40℃）を種々の基質に適用した結果をTable 3に示す。脂肪族アルデヒドを用いた場合、付加反応はスムーズに進行し、高い不斉収率で目的のプロパルギルアルコールを与えた（entries 1-7）。さらに、これまでに成功例のなかった芳香族アルデヒドを用いても反応は円滑に進行し、フェニルアセチレン（2a）をはじめとして、各種アルキンにおいて高い化学収率および不斉収率で目的物を得ることに成功した（entries 8-17）。このように各種アルデヒド及びアルキンに適用可能な幅広い基質一般性を有する触媒系の構築に成功した。

予備的な反応機構解析において、種々の光学純度を有する不斉配位子を用いて反応を行った際に、生成物の不斉収率との間に大きな正の不斉増幅が観測された（Figure 1）。丁寧な検討により、これは主にヘテロキラルな錯体の選択的な沈殿によるものであるということを明らかにした。

以上の結果は医薬合成における重要な知見であり、博士（薬学）に十分相当する研究成果と判断した。

Scheme 1. Alkynylation via Dual Activation of Both CarbonylCompounds and Alkyne

Table 1. InBr3-Catalyzed Alkynylation of Aldehydes

Table 2. In（OTf）3-Catalyzed Alkynylatioin of Ketoines

Table 3. InBr3/BINOL Complex-CatalyzedAsymmetric Alkynylation of Various Aldehydes

Figure 1. （+）-Nonlineal Effects in AsymmetricAlkynylation Catalyzed by lnBr3/BINOL Complex