Stoichiometric synthesis allylation reactions

If the alkenes and acetylenes that are subjected to the reaction mediated by 1 have a leaving group at an appropriate position, as already described in Eq. 9.16, the resulting titanacycles undergo an elimination (path A) as shown in Eq. 9.58 [36], As the resulting vinyltitaniums can be trapped by electrophiles such as aldehydes, this reaction can be viewed as an alternative to stoichiometric metallo-ene reactions via allylic lithium, magnesium, or zinc complexes (path B). Preparations of optically active N-heterocycles [103], which enabled the synthesis of (—)-a-kainic acid (Eq. 9.59) [104,105], of cross-conjugated trienes useful for the diene-transmissive Diels—Alder reaction [106], and of exocyclic bis(allene)s and cyclobutene derivatives [107] have all been reported based on this method. 

Palladium-catalyzed allylic substitution may be regarded as a special case of cross-coupling with jr-allylpalladium complexes. First developed as a stoichiometric technique, this reaction was later realized in a catalytic mode, and became a valuable tool of organic synthesis, as it allows for a broad variation of both allylic substrates and nucleophiles. 

The applications of palladium-catalyzed allylation reactions in the realm of total synthesis have been numerous, owing to the general functional group tolerance and high stereoselectivity observed in these transformations [8,17, 25, 27, 30]. In pioneering studies, Trost reported that alkene 1 was converted into allylpalladium complex 2 on treatment with a stoichiometric amounts of palladium dichloride (Scheme 14.1) [42]. Nucleophilic substitu-... 

N-Alkoxylamines 88 are a class of initiators in "living" radical polymerization (Scheme 14). A new methodology for their synthesis mediated by (TMSlsSiH has been developed. The method consists of the trapping of alkyl radicals generated in situ by stable nitroxide radicals. To accomplish this simple reaction sequence, an alkyl bromide or iodide 87 was treated with (TMSlsSiH in the presence of thermally generated f-BuO radicals. The reaction is not a radical chain process and stoichiometric quantities of the radical initiator are required. This method allows the generation of a variety of carbon-centered radicals such as primary, secondary, tertiary, benzylic, allylic, and a-carbonyl, which can be trapped with various nitroxides. 

Dialkylindolines and 1,3-dialkylindoles are formed in poor yield (<10%) from the reaction of ethyl- or phenymagnesium bromide with 2-chloro-N-methyl-N-allylaniline in the presence of catalytic quantities of (bistriphenylphosphine)nickel dichloride.72 In a modification of this procedure, the allyl derivatives can be converted by stoichiometric amounts of tetrakis(triphenylphosphine)nickel into 1,3-dialkylindoles in moderate yield72 (Scheme 43) an initial process of oxidative addition and ensuing cyclization of arylnickel intermediates is thought to occur. In contrast to the nickel system,72 it has proved possible to achieve the indole synthesis by means of catalytic quantities of palladium acetate.73 It is preferable to use... 

Although this catalytic reaction appeared to be of synthetic interest, it has since then neither been applied in synthesis nor further developed. This might be attributed in part to problems with reproducibility and catalyst stability under the reaction conditions, although the Hieber complex was used in a stoichiometric manner for the preparation of a variety of 7i-allyl-Fe complexes. These latter compounds served as starting materials for a plethora of subsequent reactions [34]. The results obtained by Nakanishi and coworkers on the stability and reactivity of n-allyl-Fe-nitrosyl complexes proved such intermediates to be reactive towards a variety of nucleophiles however, the Fe complexes formed upon nucleophilic substitution were catalytically inactive. Hence, in order to maintain the catalytic activity, the formation of intermediate 7i-allyl-Fe complexes had to be circumvented. About 3 years ago we started our research in this field and envisioned the use of a monodentate ligand to be a suitable way to stabilize the proposed catalytically active G-allyl complex. The replacement of one CO by a non-volatile basic ligand was thought to prevent the formation of the catalytically inactive 7t-allyl-Fe complex (Scheme 7.21). 

The in situ regeneration of Pd(II) from Pd(0) should not be counted as being an easy process, and the appropriate solvents, reaction conditions, and oxidants should be selected to carry out smooth catalytic reactions. In many cases, an efficient catalytic cycle is not easy to achieve, and stoichiometric reactions are tolerable only for the synthesis of rather expensive organic compounds in limited quantities. This is a serious limitation of synthetic applications of oxidation reactions involving Pd(II). However it should be pointed out that some Pd(II)-promoted reactions have been developed as commercial processes, in which supported Pd catalysts are used. For example, vinyl acetate, allyl acetate and 1,4-diacetoxy-2-butene are commercially produced by oxidative acetoxylation of ethylene, propylene and butadiene in gas or liquid phases using Pd supported on silica. It is likely that Pd(OAc)2 is generated on the surface of the catalyst by the oxidation of Pd with AcOH and 02, and reacts with alkenes. 

Several methods promoted by a stoichiometric amount of chiral Lewis acid 38 [51] or chiral Lewis bases 39 [52, 53] and 40 [53] have been developed for enantioselective indium-mediated allylation of aldehydes and ketones by the Loh group. A combination of a chiral trimethylsilyl ether derived from norpseu-doephedrine and allyltrimethylsilane is also convenient for synthesis of enan-tiopure homoallylic alcohols from ketones [54,55]. Asymmetric carbonyl addition by chirally modified allylic metal reagents, to which chiral auxiliaries are covalently bonded, is also an efficient method to obtain enantiomerically enriched homoallylic alcohols and various excellent chiral allylating agents have been developed for example, (lS,2S)-pseudoephedrine- and (lF,2F)-cyclohex-ane-1,2-diamine-derived allylsilanes [56], polymer-supported chiral allylboron reagents [57], and a bisoxazoline-modified chiral allylzinc reagent [58]. An al-lyl transfer reaction from a chiral crotyl donor opened a way to highly enantioselective and a-selective crotylation of aldehydes [59-62]. Enzymatic routes to enantioselective allylation of carbonyl compounds have still not appeared. 

In the synthesis of racemic 3-phosphoshikimic acid derivatives a palladium-catalyzed rearrangement of a cyclohexenyl phosphate was used for the stereochemical control of the C-3 oxygen function29. This rearrangement is considerably slower than the analogous reactions of allylic acetates and requires a stoichiometric amount of bis(acetonitrile)palladium(II) chloride. 

---