Tsuji-Trost allylic alkylation with ketone enolates
Tsuji-Trost allylic alkylation with ketone enolates
复制标题
DOI：
10.1002/anie.200602169
复制
发表时间：
2006-01-01
影响因子：
16.6
通讯作者：
Meier, Thorsten
中科院分区：
文献类型：
--
作者：
Braun, Manfred;Meier, Thorsten
由Tsuji首先发现并由Trost研究小组深入发展的钯催化的碳亲核试剂烯丙基取代反应是一种非常富有成效和通用的碳-碳键形成方法。通常，反应从烯丙基底物1（通常是乙酸酯或碳酸酯）开始，其在用合适的钯（0）化合物处理后形成π-烯丙基络合物2（方案1）。因此，过渡金属采用+2氧化态，并且烯丙基部分变成强碳亲电体。在随后与碳亲核试剂的反应中，在伴随的以0氧化态的贵金属的释放下获得烯丙基化产物3，从而结束催化循环。[1]已经开发了各种手性配体L*，其引导亲核试剂攻击π-烯丙基络合物2中的非对映异构末端之一，使得可以以对映选择性方式从外消旋起始材料1获得烯烃3和ent-3。此外，最近已经报道了由具有不同残基R的底物1引起的区域异构体形成问题的解决方案。[2]尽管该反应自20世纪70年代以来取得了令人印象深刻的进展，但它仍然存在一个重大缺点：碳亲核试剂类型的限制。因此，“软”碳负离子衍生的碳酸与pKa值低于20几乎被排他性地使用，丙二酸酯与二苯基烯丙基乙酸酯1（R= Ph）的组合成为一种标准程序在钯催化烯丙基化，特别是用于评估的性能手性配体L*。很明显，在这种结合中，只有一个立体中心形成在烯丙基位置。然而，如果目标是在高烯丙基位置（R216 R3，R4= H）或在烯丙基和高烯丙基位置（R216 R3，R416 H）两者中形成立体中心，如烯基酮4的逆合成合成（方案2）中所概述的，则选择的亲核试剂是预先形成的“硬”金属烯醇化物5。自20世纪70年代以来，预先形成的烯醇化物的化学和Tsuji-Trost反应同时出现和发展。[3]然而，令人惊讶的是，只有很少的尝试进行了联合收割机结合这两个概念。似乎由于早期令人失望的结果，有机化学家避免尝试进行钯催化的烯丙基化反应与预制的“硬”烯醇化物。少数几种方法并没有取得多大成功[4a]或仍然被忽视。[4b这种转化的立体选择性变体直到本世纪初才被开发出来。[5]In 1999年，Trost和Schroeder报道了第一个不稳定的酮烯醇化物的对映选择性烯丙基烷基化。[6a]因此，由2-甲基四氢萘酮（6）产生的烯醇化锡7 b进行由C2对称配体8介导的钯催化的烯丙基化（方案3）。因此，当使用导致烯丙基钯中间体从其Si面进攻烯醇化物的配体（S，S）-8时，以88%ee和99%产率获得酮（R）-9。
The palladium-catalyzed allylic substitution with carbon nucleophiles, first discovered by Tsuji and thoroughly developed by the research group of Trost, turns out to be a very fruitful and versatile method for carbon–carbon bond formations. In general, the reaction starts with an allylic substrate 1 (usually an acetate or carbonate) that, upon treatment with a suitable palladium (0) compound, forms a π-allyl complex 2 (Scheme 1). Thereby, the transition metal adopts the+ 2 oxidation state, and the allyl moiety becomes a strong carbon electrophile. In the subsequent reaction with the carbon nucleophile, the allylation products 3 are obtained under concomitant liberation of the noble metal in the oxidation state 0, thus closing the catalytic cycle.[1] Various chiral ligands L* have been developed which direct the attack of the nucleophile to one of the diastereotopic termini in the π-allyl complex 2, so that the alkenes 3 and ent-3 can be obtained from racemic starting material 1 in an enantioselective manner. Moreover, solutions to the problem of regioisomer formation arising from substrates 1 with nonidentical residues R have been reported more recently.[2] Despite the impressive progress this reaction has made since the 1970s, it has continued to suffer from a significant drawback: the limitation in the type of carbon nucleophile. Thus,“soft” carbanions derived from carbon acids with pKa values lower than 20 were used almost exclusively, and the combination of malonates with diphenylallyl acetate 1 (R= Ph) became a kind of standard procedure in palladium-catalyzed allylations, particularly for evaluating the performance of chiral ligands L*. It is obvious that, in this combination, just one stereogenic center is formed in the allylic position. If, however, the formation of stereogenic centers in the homoallylic position (R2 ¼6 R3, R4= H) or in both the allylic and the homoallylic position (R2 ¼6R3, R4 ¼6 H) is the target, as outlined in the retrosynthetic synthesis for alkenyl ketones 4 (Scheme 2), the nucleophilic reagents of choice are preformed “hard” metal enolates 5. The chemistry of preformed enolates on the one hand and the Tsuji-Trost reaction on the other hand emerged and evolved at the same time, since the 1970s.[3] Surprisingly, however, only very few attempts were made to combine both concepts. It seems that, owing to early disappointing results, organic chemists avoided attempts to carry out palladium-catalyzed allylation reactions with preformed “hard” enolates. The few approaches did not have much success [4a] or remained unnoticed.[4b, c] Stereoselective variants of this transformation, in particular, had not been developed until the beginning of this decade.[5]In 1999, the first enantioselective allylic alkylation of a nonstabilized ketone enolate was reported by Trost and Schroeder.[6a] Thus, the tin enolate 7b, generated from 2-methyl tetralone (6), was submitted to a palladium-catalyzed allylation that was mediated by the C2-symmetric ligand 8 (Scheme 3). Thereby, ketone (R)-9 was obtained in 88% ee and 99% yield when the ligand (S, S)-8, which causes the allylpalladium intermediate to attack the enolate from its Si face, was used.