Letter pubs.acs.org/OrgLett
Cite This: Org. Lett. 2018, 20, 5327−5331
Borinic Acid/Halide Co-catalyzed Semipinacol Rearrangements of 2,3-Epoxy Alcohols Kashif Tanveer, Seung-Joon Kim, and Mark S. Taylor* Department of Chemistry, University of Toronto, 80 St. George Street, Toronto, Ontario M5S 3H6, Canada
Org. Lett. 2018.20:5327-5331. Downloaded from pubs.acs.org by UNIV OF CALIFORNIA SANTA BARBARA on 09/08/18. For personal use only.
S Supporting Information *
ABSTRACT: A new mode of catalysis of the semipinacol rearrangement of 2,3-epoxy alcohols is described. In combination with a halide salt additive, diarylborinic acids promote a Type II rearrangement that occurs with net retention of configuration. This unusual stereochemical outcome is consistent with a mechanism involving regioselective ring opening of the epoxy alcohol by halide, followed by rearrangement of the resulting halohydrin-derived borinic ester. The protocol is applicable to a range of substrates, enabling ring contractions and expansions as well as stereospecific syntheses of acyclic β-hydroxycarbonyl compounds.
S
emipinacol rearrangements of 2,3-epoxy alcohols provide a useful alternative to aldol and related reactions for the synthesis of β-hydroxycarbonyl compounds. The ability to control the site of C−C bond migration (Type I1−4 versus Type II rearrangement,5−10 Scheme 1) adds to the versatility
Our group has developed catalytic transformations that hinge on the nucleophilic reactivity of tetracoordinate adducts of diarylborinic acids.16,17 We recently reported protocols for site-selective chloroacylation and chlorosulfonylation reactions of 2,3-epoxy alcohols (Scheme 2).18 Kinetics experiments and
Scheme 1. Type I and Type II Semipinacol Rearrangements of Epoxy Alcohols
Scheme 2. Borinic Acid Catalyzed Chloroacylation (Ref 18) and Semipinacol Rearrangement (This Work)
of the rearrangement approach, often enabling the construction of motifs that present challenges for aldol-based disconnections.11−14 The stereospecific nature of these rearrangements is another important attribute, given the existence of methods for the enantio- and/or diastereoselective synthesis of the 2,3-epoxy alcohol substrates.15 Herein, we describe a new mode of catalysis of the Type II semipinacol rearrangement of epoxy alcohols, employing a diarylborinic acid (Ar2BOH) in concert with a tetraalkylammonium halide salt. Unlike the majority of semipinacol and related rearrangements, this variant proceeds with net retention of configuration. Mechanistic studies suggest that the unusual stereochemical outcome is the result of a double inversion process proceeding through a borinic acid bound halohydrin diol intermediate.
computational modeling were consistent with a mechanism involving borinic acid catalyzed ring-opening of the epoxy alcohol by chloride, followed by turnover-limiting acylation or sulfonylation of the resulting tetracoordinate adduct.19 While exploring extensions of this mode of catalysis to other reagent combinations, we found that substrate 2a, upon treatment with tetrabutylammonium iodide and diphenylborinic acid (Ph2BOH, 1a), was converted to hydroxy ketone 3a, the product of a Type II semipinacol rearrangement.
© 2018 American Chemical Society
Received: July 18, 2018 Published: August 15, 2018 5327
DOI: 10.1021/acs.orglett.8b02248 Org. Lett. 2018, 20, 5327−5331
Letter
Organic Letters Several aspects of the transformation of 2a to 3a were noteworthy from our perspective. Reported semipinacol rearrangements of epoxy alcohols generally require Lewis acids that are significantly stronger than Ph2BOH and occur in the absence of Lewis basic additives, suggesting a distinct mode of activation. Moreover, the regiochemical outcome was of interest: bulky Lewis acids5,6 (often in combination with protection of the OH group to suppress chelation9) or electronically biased substrates having vinyl or aryl groups at the 3-position20 are usually required to favor Type II over the more common Type I rearrangements. Most importantly, 3a was obtained exclusively as the trans isomer: signals corresponding to the cis diastereomer21 were not observed in the 1H NMR spectrum of the unpurified reaction mixture. This stereochemical outcome is inconsistent with a concerted migration of the carbon−carbon bond to a Lewis acid activated epoxide but instead could arise from an iodohydrin diol intermediate of the type shown in Scheme 2 by a doubleinversion process. A similar mechanism was postulated more than 35 years ago by Magnusson and Thorén, who obtained cyclopentene-1-carboxaldehydes from 2,3-epoxycylohexanols by treatment with LiBr−HMPA in toluene at 110 °C.22 Because elimination to the enal products took place, the stereochemical outcome of the LiBr−HMPA-promoted rearrangement could not be determined. It should also be noted that Matsubara and co-workers observed net retention of configuration in a tandem type II semipinacol rearrangement−methylenation of epoxy alcohols promoted by the Simmons−Smith reagent bis(iodizincio)methane.8 The authors’ preliminary mechanistic proposal involved a suprafacial migration, with the Simmons−Smith reagent acting as a double Lewis acid. Catalysts and reaction conditions were evaluated for the rearrangement of substrate 2a (Table 1). For diarylborinic acid derivatives 1a−d, activity increased with the Lewis acidity of the trivalent boron center (entries 1−4), with p-fluorophenylborinic anhydride 1b providing the highest yield of 3a. (Although 1a−d are depicted in monomeric form, they exist as B−O−B dimers in the solid state. In solution, they dissociate to the active borinic acid monomers.19) Electron-deficient boronic acid 1e was inactive, as were other Lewis and Brønsted acids (entries 5−8). Both borinic acid and Bu4NI were required for the formation of 3a (entries 9 and 10). Tetrabutylammonium iodide was the optimal additive, with other salts or Lewis bases (triphenylphosphine, 1,4diazabicyclo[2.2.2]octane (DABCO)) providing inferior results (entries 11−16). Catalytic turnover of the halide salt additive was demonstrated (entry 17), but the use of a full equivalent provided a higher yield of 3a. The optimal conditions involved the use of 10 mol % of catalyst 1b and 1 equiv of Bu4NI in acetonitrile at 40 °C, giving an 80% NMR assay yield of 3a (entry 18) and 84% yield in a preparative experiment (Scheme 3). The results of a more comprehensive survey of catalysts and conditions are provided as Supporting Information (SI). The optimized protocol was applied to a collection of 12 substituted 2,3-epoxy alcohols (Scheme 3). In several of the cases examined, the readily available parent diphenylborinic acid 1a displayed similar activity to fluoro-substituted 1b. For substrates 2d and 2e, higher yields were obtained using Bu4NBr in tetrahydrofuran (THF) rather than Bu4NI in acetonitrile. The consistent formation of Type II semipinacol rearrangement products from a range of epoxy alcohol
Table 1. Evaluation of Catalysts and Additives for Semipinacol Rearrangement of Substrate 2a
entry
catalyst
additive
2aa (%)
3aa (%)
1 2 3 4 5 6 7 8 9b 10b 11 12 13 14 15 16 17b,c 18b
1a 1b 1c 1d 1e BF3.OEt2 Al(O-i-Pr)3 p-TsOH·H2O none 1b 1b 1b 1b 1b 1b 1b 1b 1b
Bu4NI Bu4NI Bu4NI Bu4NI Bu4NI Bu4NI Bu4NI Bu4NI Bu4NI none Bu4NCl Bu4NBr CsI KI PPh3 DABCO Bu4NI Bu4NI
20 10 85 90 90 95 95 85 >95 >95 85 80 50 20 80 85 25
