Rasta resin–triphenylphosphine oxides and their use as recyclable heterogeneous reagent precursors in halogenation reactions

Rasta resin–triphenylphosphine oxides and their use as recyclable heterogeneous reagent precursors in halogenation reactions Xuanshu Xia and Patrick H. Toy* Letter Address: Department of Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong, People’s Republic of China Open Access Beilstein J. Org. Chem. 2014, 10, 1397–1405. doi:10.3762/bjoc.10.143 Email: Patrick H. Toy* - Received: 13 February 2014 Accepted: 18 May 2014 Published: 20 June 2014 * Corresponding author This article is part of the Thematic Series "Organophosphorus chemistry". Keywords: Appel reaction; halogenation; organophosphorus; polymer-supported reagent; rasta resin; triphenylphosphine oxide Guest Editor: P. R. Hanson © 2014 Xia and Toy; licensee Beilstein-Institut. License and terms: see end of document. Abstract Heterogeneous polymer-supported triphenylphosphine oxides based on the rasta resin architecture have been synthesized, and applied as reagent precursors in a wide range of halogenation reactions. The rasta resin–triphenylphosphine oxides were reacted with either oxalyl chloride or oxalyl bromide to form the corresponding halophosphonium salts, and these in turn were reacted with alcohols, aldehydes, aziridines and epoxides to form halogenated products in high yields after simple purification. The polymersupported triphenylphosphine oxides formed as a byproduct during these reactions could be recovered and reused numerous times with no appreciable decrease in reactivity. Introduction One of the major drawbacks of the Wittig [1] and Mitsunobu [2,3] reactions is that they result in the formation of a stoichiometric quantity of triphenylphosphine oxide (1) as a byproduct. From an atom economy perspective this is less than ideal, and from an environmental point of view it would be good if 1 could be simply reduced to triphenylphosphine (2) for reuse [4]. In this regard Tanaka and co-workers have studied the possibility of applying the reaction first reported by Masaki and Fukui [5] in which 1 can be treated with oxalyl chloride (or bromide) to form halophosphonium salt 3a (or 3b), which in turn can be reduced to 2 under more mild reaction conditions than can 1 (Scheme 1) [6,7]. In addition to being relatively easily reduced, halophosphonium salts 3a,b are also useful reagents in a wide range of reactions, such as those illustrated in Scheme 2: (1) the conversion Scheme 1: The Masaki–Fukui reaction and halophosphonium salt reduction. 1397 Beilstein J. Org. Chem. 2014, 10, 1397–1405. Scheme 3: Catalytic Appel reactions reported by Denton and co-workers. Scheme 2: Representative reactions involving halophosphonium salts 3a,b. of alcohols 4 to alkyl halides 5 (the Appel reaction), (2) the conversion of aldehydes 6 to 1,1-dihaloalkanes 7, (3) halogenation of aziridines 8 to form 2-haloamines 9, (4) halogenation of epoxides 10 to form 1,2-dihaloalkanes 11, (5) and the dehydration of oximes 12 to form nitriles 13. Capitalizing on the fact that 1 is formed as a byproduct from 3a,b in each of these reactions, Denton and co-workers have recently combined the Masaki–Fukui reaction with many of the functional group transformation outlined in Scheme 2, in onepot processes in which the role of 1 is referred to as that of a catalyst [8-12]. For example, catalytic Appel reactions were achieved by slowly adding separate solutions of oxalyl chloride and alcohols 4 to a solution of 1 (Scheme 3) [8,9]. In these reactions, the simultaneous slow addition of oxalyl chloride and alcohol substrate 4 to a sub-stoichiometric quantity of 3a was necessary in order to minimize formation of undesired ester side-products formed by the reaction of 4 with the acid chloride. Furthermore, chromatographic purification of the alkyl halide product 5 was required. Thus, while the procedures reported by Denton et al. might be conceptually interesting, they may not be particularly convenient to perform, especially on larger scales than what was originally reported. We have had a long-term interest in the use of organic polymers as supports for reagents and catalysts [13], and have reported the use of various polymer-supported phosphines as reagents, organocatalysts, and ligands in order to simplify product isolation [14-18]. Most recently we have studied the use of the rasta resin polystyrene architecture [19-26] as a platform for reagents and catalysts [27-33], and have used easily synthe- sized rasta resin–Ph3P (14) in various Wittig reactions that required only filtration and solvent removal for product purification (Figure 1) [27-29]. Additionally, 14 was converted into phosphonium salt 15, which proved to be an efficient and highly recyclable catalyst for aldehyde and ketone cyanosilylation reactions from which the products could also be obtained pure after only filtration and solvent removal [30]. It should be noted that the grafts of the rasta resins reported are random co-polymers, and the structures drawn for them are not mean to indicate that they are block co-polymers. The format for their presentation is used merely to indicate their monomer incorporation ratios. Figure 1: Rasta resins 14 and 15. Thus, considering our prior success in using 14 and 15, we wanted to oxidize 14 to 16, and in turn use this as a heterogeneous precursor to reagents 17a,b for use in the halogenation reactions described in Scheme 2. We anticipated that using a full equivalent 17a,b generated in situ would eliminate the need for slow addition of the oxalyl halide to form the halophosphonium salt, and thus allow for the reactions to be performed more conveniently than in the catalytic procedures of Denton and 1398 Beilstein J. Org. Chem. 2014, 10, 1397–1405. co-workers. Furthermore, since 16 would be the byproduct of the reactions, it could be recovered by filtration at the end of the reactions and reused directly. Herein we report the realization of this strategy and describe simple procedures for alcohol, aziridine, aldehyde and epoxide halogenation reactions from which the desired products are easily isolated and the phosphine oxide byproduct is readily recycled. Results and Discussion Rasta resin 16 was prepared by oxidation of 14, which was prepared as previously reported [28], using H2O2 (Scheme 4). The loading level of 16 was determined by elemental analysis to be 0.97 mmol/g, and gel-phase 31P NMR spectroscopic analysis of 16 showed only a single peak at 29.4 ppm, indicating that the phosphine groups were completely oxidized. Appel reactions using 16 With 16 in hand, we initially used it to perform Appel reactions by first converting it into either 17a or 17b in situ (Scheme 4). To do this, 16 was suspended/swollen in dichloromethane, and then the appropriate oxalyl halide was added. Once gas evolution ceased, alcohol 4 was added, and the reaction mixture was heated to reflux. Progress of the reactions was monitored by TLC analysis, and they were all finished in 4–7 hours. Upon completion, the reaction m (...truncated)