Catalytic Diastereoselective Reductive Aldol Reaction: Optimization of Interdependent Reaction Variables by Arrayed Catalyst Evaluation
Catalytic Diastereoselective Reductive Aldol Reaction: Optimization of Interdependent Reaction Variables by Arrayed Catalyst Evaluation
复制标题
DOI：
10.1021/ja992952e
复制
发表时间：
1999-12
影响因子：
15
通讯作者：
Steven J. Taylor and;J. Morken
中科院分区：
文献类型：
--
作者：
Steven J. Taylor and;J. Morken
介绍温和的，立体选择性的，和催化方法合成聚丙酸酯是当前感兴趣的主题。虽然这种键形成的大多数方法采用甲硅烷基烯醇醚和刘易斯酸催化剂，1报道的Co-，Rh-，Pt-和Pd-催化的丙烯酸酯，醛和硅烷之间的缩合反应（方程式1）也显示出合成羟醛加合物的前景。2-4这种还原性羟醛缩合反应的一个优点是不需要活化烯醇化物的化学计量预形成。虽然只有很少的文献先例描述R，β-不饱和酯和醛的还原羟醛偶联，但从这些报道中显而易见的是，可以使用各种后过渡金属催化剂。虽然这些反应通常实现有用的产物产率，但非对映选择性仍然具有挑战性（最大4：1顺式：反式选择性）。关于不对称催化反应的研究还很少，反应机理也知之甚少。本文中，我们公开了从192个独立催化体系的高通量评价中获得的用于立体选择性还原性羟醛缩合反应的有效催化剂体系的发现。5，6除了揭示具有合成效用的催化剂之外，这些研究还说明了反应变量的显著相互依赖性，从而为未来的催化剂开发预先假定了阵列催化剂评估。7从一开始，我们就预期过渡金属盐、配体和氢化物源将影响催化还原性羟醛缩合反应中的反应性和选择性。我们还预计，在缺乏大量机械数据的情况下，正确选择每个变量将具有挑战性。因此，我们选择评价这些反应组分的多种不同组合（图1）。为了检查上述反应变量的影响，我们在玻璃96孔板中进行了一系列实验。在我们最初的阵列中，我们使用了四种过渡金属盐，七种配体（加上空白）和六种氢化物源。金属和金属化物包括已知用于催化烯烃还原的那些。选择8个配体9以实现最大的官能团多样性。在实验中，金属和配体在50 ℃下在二氯乙烷中预混合1小时。在室温下将每种催化剂与氢化物试剂一起温育30分钟后，加入苯甲醛和丙烯酸甲酯（20：1底物：催化剂），并使反应在室温下进行16小时。在酸性处理后，通过手性GC相对于内标物分析每个反应。以这种方式，确定每个实验的相对转化率和立体异构体比率。图1显示了上述192个独立实验10中每一个的相对产率，并揭示了反应条件和产率之间的许多值得注意的关系。首先，邻苯二酚硼烷倾向于与最多数量的催化剂发生反应，而Cl 3 SiH仅在MOP配体存在下与[（烯丙基）-PdCl] 2发生反应。第二，当用一种氢化物源取代另一种氢化物源时，反应性特征通常相反：当使用Et 2 MeSiH时，[（cod）IrCl] 2因添加Ph-半卟啉配体而中毒（无配体时78%相对产率，有配体时0%相对产率），尽管当使用PhSiH 3时，相同的金属盐被Ph-半卟啉活化（无配体时2%相对产率，有配体时24%相对产率）。反应变量的这种相互依赖性反映在观察到三种最活跃的催化剂体系（[（cod）-RhCl] 2-DuPhos-Cl 2 MeSiH、Co（acac）2-MOP-PhSiH 3和[（cod）-RhCl] 2-binap-catechol borane）中没有一种是...
Introduction of mild, stereoselective, and catalytic processes for the synthesis of polypropionates is a topic of current interest. While most approaches to such bond formation employ silyl enol ethers and Lewis acid catalysts, 1 reports of Co-, Rh-, Pt-, and Pd-catalyzed condensation between acrylate esters, aldehydes, and silanes (eq 1) have also shown promise for the synthesis of aldol adducts. 2-4 One advantage of such a reductive aldol reaction is that stoichiometric preformation of an activated enolate is not required. While there is only scant literature precedent describing the reductive aldol coupling of R, β-unsaturated esters and aldehydes, it is apparent from these reports that a variety of late transition metal catalysts may be used. Although useful product yields are often realized for these reactions, diastereoselection remains challenging (maximum 4: 1 syn: anti selectivity). No efforts have been made in regards to asymmetric catalysis and little is known about the reaction mechanism. Herein, we disclose the discovery of an effective catalyst system for the stereoselective reductive aldol reaction obtained from high-throughput evaluation of 192 independent catalytic systems. 5, 6 In addition to revealing a catalyst with synthetic utility, these studies illustrate a remarkable interdependence of reaction variables and thereby presuppose arrayed catalyst evaluation for future catalyst development. 7 From the outset, we expected that the transition metal salt, ligand, and hydride source would affect reactivity and selectivity in the catalytic reductive aldol reaction. We also expected that, in the absence of substantial mechanistic data, proper choice of each variable would be challenging. Therefore, we chose to evaluate a number of different combinations of these reaction components (Figure 1). To examine the effect of the abovementioned reaction variables, we performed an array of experiments in glass 96-well plates. In our initial array, we employed four transition metal salts, seven ligands (plus a blank), and six hydride sources. The metals and hydrides included those known for catalytic alkene reduction. 8 Ligands9 were chosen to achieve the greatest functional group diversity. In the experiment, the metals and ligands were premixed at 50 C in dichloroethane for 1 h. After incubating each catalyst with the hydride reagent for 30 min at room temperature, benzaldehyde and methyl acrylate (20: 1 substrate: catalyst) were added and the reaction was allowed to proceed at room temperature for 16 h. After acidic workup, each reaction was analyzed by chiral GC versus an internal standard. In this manner, relative conversion and stereoisomer ratios were determined for every experiment. Figure 1 shows the relative yield for each of the 192 independent experiments described above10 and reveals a number of noteworthy relationships between reaction conditions and yield. First, catechol borane tends to give reaction with the largest number of catalysts whereas Cl3SiH is effective only with [(allyl)-PdCl] 2 in the presence of MOP ligand. Second, reactivity characteristics are often opposed when substituting one hydride source for another:[(cod) IrCl] 2 is poisoned by the addition of Ph-semicorrin ligand when Et2MeSiH is used (78% relative yield without ligand, 0% relative yield with ligand) although the same metal salt is activated by Ph-semicorrin when PhSiH3 is used (2% relative yield without ligand, 24% relative yield with ligand). This interdependence of reaction variables is reflected in the observation that none of the three most active catalyst systems ([(cod)-RhCl] 2-DuPhos-Cl2MeSiH, Co (acac) 2-MOP-PhSiH3, and [(cod)-RhCl] 2-binap-catechol borane) are …