Phosphonates Perkow reaction

Protocol 2 produces the protected fi-formylphosphonate 12 f)-ketophospho-nates may also be synthesized by other methods,23 however, they may not be prepared in unprotected form by the Michaelis-Arbuzov reaction because the Perkow reaction, in which an a-haloaldehyde or ketone and a trialkyl phosphite yield an enol phosphate (e.g. 13, Scheme 5,24 i.e. [P—O] bond formation), competes and frequently dominates (see Section 4). Conversely halocarboxylic acid derivatives (e.g. see Table 7.1, entry 3) and acyl halides (see Protocol 3) react well in the Michaelis-Arbuzov reaction to yield useful functionalized phosphonates. fi-Ketophosphonates are useful reagents for the synthesis of a,fi-unsaturated carbonyl compounds by the Horner-Wadsworth-Emmons reaction,3,4 25 and have other applications.23... 

Dialkyl phosphites and their metal salts seldom undergo the Perkow reaction with a-halo-aldehydes or -ketones but usually yield a-hydroxy- and/or epoxy-phosphonate esters (i.e. Pudovik reaction products, cf. Section 6).78... 

Possibly the most frequently used and most widely known phosphonylated aldehyde is diethyl l-formylmethyIphosphonate. A vaiiety of methods for the preparation of this aldehyde are reported in the literature. The oldest employs tlie Michaelis-Arbuzov reaction between triethyl phosphite and bromoacetaldehyde diethyl acetal, which yields diethyl 2,2-diethoxyethylphosphonate on heating to 160°C (Scheme 5.5). Subsequent acid hydrolysis gives diethyl 1-formylmethylphosphonate. The aldehyde function requires protection because it is known that a-haloaldehydcs react with trialkyl phosphites in a Perkow reaction affording dialkyl vinylphosphates isomeric with the expected phosphonates. " ... 

In all cases mentioned at that time, the products were assumed to be phosphonates of the type expected from a normal Michaelis-Arbuzov reaction. - However, in light of the Perkow reaction, all the P-ketophosphonates described in numerous papers and patents have to be formulated as the isomeric vinyl phosphates. 

In 1952 Perkow discovered a new type of rearrangement in which trialkyl phosphites react with a-aldehyde.6,7 This anomalous Arbuzov reaction, which resulted in the formation of dialkylvinylphosphates, subsequently became known as the Perkow reaction. 1 The reaction of trialkyl phosphites with alkyl halides (Arbuzov reaction) proceeded to yield dialkoxy alkyl phosphonates.8 The alkyl halide undergoes a nucleophilic displacement to yield the nonisolable alkyl trialkoxy phophonium halide.8 The halide can then attack an alkoxy group to give phosphonate ester and RX. An increase in temperature favors the Arbuzov reaction leading to phosphonates (Sn2 displacement). 

Many of the reactions applied to the synthesis of phosphonoyl and phosphinoyl alkanoic acids described in the previous section can also, in principle, be applied to the synthesis of oxoalkyl-phosphonic and -phosphinic acids. A notable exception, however, is the synthesis of those compounds in which the oxo group is in the j5-position relative to phosphorus, when important syntheses lead not to phosphonate or phosphinate esters, but rather to enol esters of phosphoric acid through what is now referred to as the Perkow reaction ... 

Steric hindrance at the carbonyl group restricts the Perkow reaction and facilitates direct Michaelis-Arbuzov displacement of the halogen to result in increased yields of the (2-oxoalkyl)phosphonate thus, 2,4,6-trimethylphenylacetyl halides yield only the... 

Nearly 50 years after the initial reports of Arbuzov and Michaelis, and despite early reports affirming the contrary, it was found that a-haloaldehydes (and, later, some a-haloketones and a-haloesters) underwent reaction with phosphites to produce vinylphosphates rather than the expected phosphonates.The path (Scheme 10.51, the Perkow reaction) is presumed to involve attack of phosphorus on the carbon of the carbonyl and subsequent rearrangement. 

The preparation of carbohydrate enol phosphates, using the Perkow reaction, and of uridine 2, 3 -(adamant-l-yI)phosphonates are dealt with in Chapters 18 and 21, respectively. 

To obtain more information on the nature of the quasiphosphonium intermediates involved in these systems we have studied the reactions gf sterically hindered neopentyl esters by means of 1P nmr spectroscopy. Trineopentyl phosphite and a-bromoacetophenone gave rise to a peak at +41 ppm due to the ketophosphonium intermediate 3 (R = Me.CCH, R = Ph X = Br ) within half an hour of mixingJthe reactants in acetone-dfi at 27 °C ( p nmr shifts are relative to 85% H-PO. down-field positive). Peaks due to the ketophospnonate 4 +19 ppm and the vinyl phosphate 7 (-7 ppm) were also observed (compound 4 and 7 have satisfactory elemental analysis and spectroscopic data ). The concentration of the intermediate reached a maximum after about two hours when it was precipitated from acetone solution by the addition of anhydrous ether to give white crystals of trineopentyloxy (phenacyl)phosphonium bromide, identified by elemental analysis and nmr spectroscopy ( XP 6+41, in CDCl ). When redissolved in acetone-dg, deuterochloroform, acetic acid, or acetic acid-acetone mixtures, the intermediate decomposed to yield keto-phosphonate 4 but none of the vinyl phosphate 6 (Perkow product). Nor was the course of reaction affected by the addition of chloride ion or of a-chloro-acetophenone in acetonitrile. 

Between 1948 and 1952 Morris and Van Winkle (1952), Arbuzov and Alimov (1951), and Perkow et al. (1952) investigated independently the reaction of polyhalogen aldehydes, primarily of chloral, with trialkyl phosphites. The conversion was assumed by some workers to be a true Michaelis-Arbuzov reaction and, accordingly, it was thought erroneously that the product formed in the reaction of chloral and trimethyl phosphite is the phosphonate of formula 66. Perkow et al. (1952) established that a-halogen carbonyl compounds give with trialkyl phosphites an anomalous Michaelis-Arbuzov reaction, so that the actual reaction product is 2,2-dichlorovinyl phosphate, 67. This reaction is known in the literature as Perkow s reaction. 

In 1952, Perkow reported (242) that a-haloaldehydes did not react with trialkyl phosphites according to the Michaelis-Arbuzov reaction, although this had been repeatedly reported, but that a new type of reaction occurred, yielding dialkyl vinyl phosphates isomeric with the expected phosphonates. 

---