On the mechanism of catalytic, enantioselective allylation of aldehydes with chlorosilanes and chiral Lewis bases
On the mechanism of catalytic, enantioselective allylation of aldehydes with chlorosilanes and chiral Lewis bases
复制标题
DOI：
10.1021/ja002060a
复制
发表时间：
2000-12-06
影响因子：
15
通讯作者：
Fu, JP
中科院分区：
文献类型：
--
作者：
Denmark, SE;Fu, JP
烯丙基金属试剂与醛的对映选择性加成反应是一种有效的、通用的立体选择性碳-碳键形成方法。1近年来出现的一个更有用的变体是路易斯碱促进的烯丙基和巴豆基三卤代硅烷的加成反应。2，3 1994年，第一例用手性磷酰胺与烯丙基三氯硅烷的对映选择性加成反应在这些实验室中被披露，方案1。从那时起，一些基团报道了手性磷酰胺、5a、b甲酰胺、5c、d和N-氧化物、5e脲、5f和二胺促进的对映选择性加成反应。5G尽管通过对启动子结构的经验修饰来提高对映体选择性做出了很大努力，但对于速率加速和立体选择的起源仍然缺乏明确的机制图景。我们对三氯硅烯醇酸盐相关反应的研究揭示了不同的反应途径，包括一级和二级对催化剂的依赖以及阳离子氯硅酸盐物种的中间体。6我们现在提供了动力学、立体化学和结构证据，支持在烯丙化过程中运行类似的途径。在我们的初步披露4中，我们注意到，在降低催化剂负载量的情况下，尽管在反应条件下缺乏竞争的、非催化的成分，但对映体选择性受到侵蚀。7这表明，反应可能通过两个磷酰胺与氯硅烷结合的途径以及只涉及一个催化剂分子的选择性较低的途径进行。为了获得这一假说的支持，我们利用了卡根开创的强大的方法，即通过非线性效应进行不对称放大。8这项研究的结果，如图1所示，清楚地显示了适度的(G)(0.46)，但真实的，积极的非线性效应。9观察到的不对称扩增被解释为在立体化学决定的过渡结构中存在两个(R，R)-3分子。为了确定这两个磷酰胺分子是否也存在于速率决定步骤中，我们确定了每个组分的总速率表达式和顺序。为此，通过使用ReactIR 1000型仪器对苯甲醛的消耗进行现场监测，确定了烯丙化反应的动力学参数。用1a(10当量)和1当量的(R，R)-3的大过量建立了苯甲醛的11级。绘制-ln[苯甲醛]与时间的直线(R2)为0.9984，从而建立了醛的一级依赖关系。通过测定等摩尔浓度下的总反应级数，间接建立了1a中的12a级。在本实验中，[苯甲醛]-1与时间的关系图给出了一条直线(R2)0.9986，表明反应总体上是二级反应12b，因此是1a的一级反应。通过测定不同促进剂浓度(-78℃)下的动力学速率常数，确定了在磷酰胺中的反应级数。在这些实验中，使用等摩尔量的1a和苯甲醛，催化剂的负载量为50-400mol%。二阶速率常数(ln(Kobs))与催化剂浓度(ln[(R，R)-3])的ln/ln图为直线(R2)0.9987，斜率为1.77.13很明显，该反应对催化剂表现出较高的级数依赖性。该命令的原因是
The enantioselective addition of allylmetal reagents to aldehydes is now well established as a powerful and general method for stereoselective carbon-carbon bond formation. 1 One of the more useful variants to emerge in recent years is the Lewis-base promoted addition of allyl-and crotyltrihalosilanes. 2, 3 In 1994, the first example of enantioselective addition of allylic trichlorosilanes by the use of chiral phosphoramides was disclosed from these laboratories, Scheme 1. 4 Since then, a number of groups have reported enantioselective additions promoted by chiral phosphoramides, 5a, b formamides, 5c, d and N-oxides, 5e ureas, 5f and diamines. 5g Despite significant efforts at improving the enantioselectivity by empirical modification of the promoter structure, a clear mechanistic picture for the origin of rate acceleration and stereoselection is still lacking. Our ongoing investigations on the related reactions of trichlorosilyl enolates have revealed divergent pathways involving both first-and second-order dependence on catalyst and the intermediacy of cationic chlorosiliconate species. 6 We now provide kinetic, stereochemical, and structural evidence in support of a similar pathway operating in the allylation process.In our preliminary disclosure4 we noted that at reduced catalyst loadings the enantioselectivity eroded despite the lack of a competitive, uncatalyzed component under the reaction conditions. 7 This suggested the possibility that a the reaction could proceed by a pathway involving two phosphoramides bound to the chlorosilane along with a less selective pathway involving only one catalyst molecule. To gain support for this hypothesis, we made use of the powerful method, pioneered by Kagan, of asymmetric amplification by nonlinear effects. 8 The results of this study, graphically depicted in Figure 1, clearly demonstrate a modest (g) 0.46), but real, positive nonlinear effect. 9 The observed asymmetric amplification is interpreted as arising from the presence of two molecules of (R, R)-3 in the stereochemically determining transition structure. 10 To establish if the both phosphoramide molecules were also present in the rate-determining step, we determined the overall rate expression and order in each component. Toward that end the kinetic parameters of the allylation were determined by in situ monitoring of the consumption of benzaldehyde by the use of a ReactIR 1000 instrument. 11 Order in benzaldehyde was established by using a large excess of 1a (10 equiv) and 1 equiv of (R, R)-3. Plotting-ln [benzaldehyde] versus time gave a straight line (R2) 0.9984), thus establishing first-order dependence in aldehyde. 12a Order in 1a was established indirectly by determining the overall reaction order at equimolar concentration. For this experiment, a plot of [benzaldehyde]-1 versus time gave a straight line (R2) 0.9986), indicating that the reaction is overall second order12b and therefore first order in 1a. The reaction order in phosphoramide was established by determining the kinetic rate constants at various promoter concentrations (at-78 C). For these experiments equimolar amounts of 1a and benzaldehyde were used at catalyst loadings of 50-400 mol%. A ln/ln plot of the second-order rate constants (ln (kobs)) versus the catalyst concentration (ln [(R, R)-3]) gave a straight line (R2) 0.9987) with a slope of 1.77. 13 Clearly the reaction displays a higher-order dependence on catalyst. The reason that the order