Catalytic Wittig and aza-Wittig reactions

Catalytic Wittig and aza-Wittig reactions Zhiqi Lao and Patrick H. Toy* Review Address: Department of Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong, People’s Republic of China Open Access Beilstein J. Org. Chem. 2016, 12, 2577–2587. doi:10.3762/bjoc.12.253 Email: Patrick H. Toy* - Received: 29 September 2016 Accepted: 14 November 2016 Published: 30 November 2016 * Corresponding author This article is part of the Thematic Series "Green chemistry". Keywords: aza-Wittig reactions; catalysis; phosphines; phosphine oxides; reduction; silanes; Wittig reactions Guest Editor: L. Vaccaro © 2016 Lao and Toy; licensee Beilstein-Institut. License and terms: see end of document. Abstract This review surveys the literature regarding the development of catalytic versions of the Wittig and aza-Wittig reactions. The first section summarizes how arsenic and tellurium-based catalytic Wittig-type reaction systems were developed first due to the relatively easy reduction of the oxides involved. This is followed by a presentation of the current state of the art regarding phosphine-catalyzed Wittig reactions. The second section covers the field of related catalytic aza-Wittig reactions that are catalyzed by both phosphine oxides and phosphines. Introduction The Wittig reaction is a venerable transformation for converting the carbon–oxygen double bond of an aldehyde or a ketone into a carbon–carbon double bond of an alkene group (Scheme 1). Since its introduction over half a century ago [1,2], it has been widely employed in organic synthesis due to its versatility and reliability. The requirement of simple and inexpensive reagents to generate the necessary phosphonium ylide (phosphorane) reactant (a phosphine, typically Ph3P (1), an alkyl halide and a base), also adds to its appeal [3,4]. However, despite its proven utility, the Wittig reaction suffers from limitations that may deter from its use, especially on a large scale, in the context of green sustainable chemistry. For example, it has low atom economy due to its requirement of one molar equivalent of a phosphine reagent, and the formation of a corresponding amount of a phosphine oxide, usually Ph3P=O (2). There is also the associated problem of separating a by-product from the desired product when they are formed in equal molar amounts. These major deficiencies of the Wittig reaction have led to numerous efforts towards developing variations of it which are catalytic in the required phosphine, or a surrogate for it, and this research is the major focus of this review [5-8]. Additionally, analogous catalytic aza-Wittig reactions, in which carbon–nitrogen double bonds of imine groups are formed, will also be discussed in the second section of this review. 2577 Beilstein J. Org. Chem. 2016, 12, 2577–2587. phite to regenerate catalyst 3 for participation in another reaction cycle. Overall, the reaction conditions were quite mild, with the reactions being performed at room temperature with only slight excesses of base and reducing reagent being required. It should be noted that the use of only electron-withdrawing groups activated alkyl halides 4, and that aromatic and aliphatic aldehydes 6 worked well in these reactions to produce products 7 in high yields with predominantly E-configuration. Scheme 1: Prototypical Wittig reaction involving in situ phosphonium salt and phosphonium ylide formation. Review Catalytic Wittig reactions A key requirement for versions of the Wittig reaction that are catalytic in phosphine is the selective in situ reduction of the P(V) phosphine oxide byproduct back to the P(III) phosphine in the presence of a reducible aldehyde or ketone substrate, an alkyl halide and a base. Thus, it seems that the challenge in developing catalytic versions of the Wittig reaction distils down to identifying and implementing selective reducing conditions that enables the necessary catalyst redox cycling, yet does not reduce either the starting materials or the desired alkene-containing product. Quite a few years later Yong Tang and co-workers, also of the Shanghai Institute of Organic Chemistry, carried on with this research and extended it by using a combination of Ph 3 As (9, 0.2 equivalents), Fe(TCP)Cl (10, TCP = tetra(p-chlorophenyl)porphyrinate), and ethyl diazoacetate (11) to generate arsonium ylide 12 for use in biphasic catalytic Wittig-type reactions (Scheme 3) [11]. In these reactions sodium hydrosulfite replaced triphenylphosphite as the reducing reagent to convert the byproduct Ph3As=O (13) back into 9 in the aqueous phase of the reaction mixture in order to make the reactions more environmentally friendly. As was the case in the previous work described above, both aromatic and aliphatic aldehydes 6 were suitable substrates in this reaction system to form products 7. Arsine and telluride-catalyzed reactions As phosphine oxides are generally very stable and relatively difficult to reduce, the group of Yao-Zeng Huang used their prior findings that arsonium ylides can participate in Wittigtype reactions. Further they found that arsine oxides can be reduced using much milder reaction conditions compared to phosphine oxides. They developed the first reported catalytic Wittig-type reactions in which Bu3As (3, 0.2 equivalents) was used as the catalyst (Scheme 2) [9,10]. The reaction of 3 with an alkyl halide 4 followed by deprotonation using potassium carbonate generated the corresponding arsonium ylide (5) which, in turn, reacted with an aldehyde substrate 6 to produce the alkene-containing product 7 together with Bu3As=O (8). The byproduct 8 was then reduced in situ using triphenylphos- Scheme 3: Ph3As-catalyzed Wittig-type reactions using Fe(TCP)Cl and ethyl diazoacetate for arsonium ylide generation. Scheme 2: Bu3As-catalyzed Wittig-type reactions. 2578 Beilstein J. Org. Chem. 2016, 12, 2577–2587. Most recently the Tang research group has reported the use of polymer-supported arsine 14 (0.008–0.04 equivalents) as the catalyst in related reactions (Figure 1) [12]. In this work, 14 was found to be the only arsine examined that was able to effectively catalyze Wittig-type reactions of ketone substrates to produce tri- and tetrasubstituted alkene products in very high yields. For these reactions, which required a higher operating temperature than before (110 °C compared to 80 °C), polymethylhydrosiloxane was used as the reductant, and 14 could be recovered and reused efficiently in numerous reaction cycles without loss of catalytic activity. Tang’s research group also followed up this tellurium-based research many years later and published several papers describing the use of polymer-supported tellurides, such as 18, as catalysts (Scheme 5) [14-16]. The major advantage reported for using 18 instead of 15 is that a much lower catalyst loading could be used in similar reactions (0.02 equivalents compared to 0.2 equivalents). Unfortunately, despite the fact that 19 could be easily re (...truncated)